Guest post by Why am I writing this?
After reading reviews from ArsTechnica and phoronix.com about OpenSolaris, here is what I know about OpenSolaris based on those two reviews:
1.That OpenSolaris uses the LiveCD format similar to Ubuntu and Knoppix.
2.That the OpenSolaris community has improved the installer to make a more “friendly” installation experience.
3.That the Gnome desktop environment is used and includes the standard assortment of desktop applications, except for office software (OpenOffice).
4.That OpenSolaris doesn’t ship with the latest and greatest applications.
5.That hardware support is still a problem.
6.That Sun has a long way to go before the reviewers are happy with OpenSolaris.
What I did not find in either article:
1.Any significant attention paid to ZFS as the default filesystem other than the mention of it in the ArsTechnica piece.
2.What network hardware was used by Ryan Paul in the ArsTechnica article that he had trouble with.
3.That bash replaced the Bourne shell as the shell for root and ksh was upgraded to ksh93.
4.What is included under /usr/gnu.
5.What did not ship with OpenSolaris that would normally be associated with Solaris or Solaris Express.
6.No mention of root as a role or how to change it.
While the phoronix review was conducted with a preview release of OpenSolaris, the ArsTechnica article was published a week after the release. Neither review goes into much depth beyond the GUI, installed applications and a list of wishes and complaints. While this might be acceptable to some, I was hoping for something more than a quick look to see if OpenSolaris is in parity with Ubuntu or Fedora as a desktop OS.
Starting OpenSolaris
Using the Live CD has one distinct benefit and that is the Device Driver Utility, running this tool identifies the hardware in your machine and shows you where you could have problems in terms of device support based on the drivers that ship with OpenSolaris. If your hardware is not supported than all you have wasted is the time to download OpenSolaris and burn it to a CD.
Installation
I had minor difficulties when attempting to install OpenSolaris on my Gateway 7405 GX laptop that had a NTFS partition where if I selected “Use whole disk” the installation would fail. Selecting partition the disk allowed me to proceed with the installation. The installation routine starts off with two selections about language that are text screens, once those selections are made the graphical installer runs and the rest of the installation is performed through the installer.
While some are happy with the progress that Sun has made with the installation routine, others still complain that the first part of the installation routine is text based. I do not see a problem with the installation routine at all. This is not just because I have used Sun products for years and have become used to the install routine. The install program is simple and straightforward, I don’t see the need to improve something that works.
Bash as a root shell
One the major changes from Solaris and Solaris Express is using bash as the shell for the root user as opposed to the Bourne shell, which has been a part of Solaris since version 1. I do not spend that much time as root to justify a shell with an increased level of usability since the vast majority of the administration I perform is at the sysadmin group level (14). For those who want to change the shell back to /bin/sh, or whatever you like is a two step process, the first is to remove root as a role using the following command:
rolemod -K type=normal root
Once that is done, using usermod -s, you can change the shell for root to the desired shell.
The /usr/gnu directory tree
Checking under the hood one finds /usr/gnu, where GNU specific utilities can be found. The complaints I mostly hear are for gtar (/usr/bin/gtar), autoconf which is not part of the distribution or available at pkg.opensolaris.org, and neither is automake, but that can be downloaded from pkg.opensolaris.org. For some software that is written specifically to be compiled under Linux or with the complete GNU tool chain, there is potential for problems since the tool chain might not be complete enough depending on what you are trying to compile.
Getting On the Net and configuring interfaces
One of the many things people complained about, including Ryan Paul of ArsTechnica is the difficulty in getting an OpenSolaris machine on the network and to access the Internet. For the purposes of writing this review I used two machines to compare and contrast setting up networking. The first one is a home built Pentium IV machine using an Asus motherboard with 1.5 GB of memory, 2 250 GB Western Digital IDE disks, an ATI Radeon 9100 AGP card, and a Intel Pro/1000 PCI NIC. The second machine is a Gateway laptop with an Athlon 64 CPU, 1.25 GB of memory, a 60 GB 7200 RPM hard disk, a ATI Mobility Radeon 9600 video card, a VIA Rhine II 10/100 NIC, a Broadcom internal wireless card and a Proxim Orinoco Gold PCMCIA wireless card.
Using the Pentium IV rig nwamd set up my Intel Pro/1000 NIC to an IP address of 192.168.1.180 using the DHCP server OpenSolaris runs (at the time I installed OpenSolaris on this machine I did not use DHCP on my router), the entries for DNS were wrong, and as a result I had only local network connectivity. I modified the /etc/resolv.conf to populate the file with good DNS servers and modified the hosts entry of /etc/nsswitch.conf and was immediately able to access the Internet.
The Gateway laptop’s internal Broadcom wireless and the VIA Rhine II 10/100 interfaces had no drivers, but the Proxim Orinoco Gold PCMCIA card was recognized by OpenSolaris and ready for use, discovered my wireless access point and I was prompted for my WEP key. Once that was supplied the machine received an IP address from my router after starting the DHCP server and my laptop was on the network.
Using ips
When I wanted to compile iozone to run some benchmark tests against ZFS volumes, I discovered that gcc was not included as part of the OpenSolaris LiveCD. The standard answer for a Solaris user would be to use either Steve Christainsen’s SunFreeware or Dennis Clarke’s BlastWave to get gcc, with OpenSolaris there is the Image Packaging System which is new to OpenSolaris. Checking with the web site pkg.opensolaris.org I found gcc was available and to download and install the package took one command from a term window:
pkg install SUNWgcc
The package was downloaded and installed successfully, and running which gcc produced the result of /usr/bin/gcc. For those who prefer a graphical tool, a package manager is included as part of the desktop. While there are some people who think a graphical package management tool is a necessity, Michael Larabel states “and expecting some users to use the command-line to install packages is not an option”. I don’t agree with that assessment at all, I see it as a nice to have since the command line version works from either a term window or the console if you have Gnome shut off. It simply comes down to a matter of personal preference and what you are trying to achieve.
While ips may appease Linux users for now, the bigger question is how long will it be before the standard complaint about Sun’s repository not having the latest software comes up. That question was answered by both Ryan and Michael in their respective articles where they pointed out that Gnome was not as up to date as Ubuntu and Fedora. The real question is whether the difference between the versions provide so much increased functionality that it would be a show stopper for users, or is it just more complaining?
Sun tends to be a little conservative when it comes to adding software to Solaris, and I don’t see the trend changing with OpenSolaris. I see this as good because I would rather have a stable platform that works than one that has all of the latest software that doesn’t work.
ZFS as the default filesystem
The biggest, and, for many, the most anticipated change is using ZFS as the default filesystem over UFS, the current standard for Solaris. This is the feature I have been waiting for Sun to put into Solaris for some time. I recently had to replace a bad disk in a ZFS mirror on one of my production systems at work, and there is nothing I like more than detaching the bad disk with a single command, removing the bad disk from the system and replacing it with the new disk and attaching it with a single command. Anyone having experience with Solaris Volume Manager or Solstice Disk Suite will quickly appreciate what ZFS brings to the table in terms of ease of administration of volumes.
ZFS performance
A lot has been said about the performance of ZFS, in many cases people have said the performance of ZFS is poor, particularly on 32-bit hardware without much in the way of data to back it up. I decided to test for myself what the performance of ZFS was for the two systems I am using for the review using iozone by William Norcott and Don Capps. In both cases iozone was compiled as a 32-bit binary and the following options were used at runtime, -RAb machine name.wks -g 2G. I used the -A option for small file sizes as well as the larger ones normally tested by iozone. The 2G option specifies testing up to a 2 gigabyte file size, so that any disk caching is no longer effective, thus what is being measured at the point the cache is exhausted is the performance of the raw disk or disks.
The Pentium IV machine is configured with a ZFS rpool mirror using two Western Digital 250 GB disks across the same IDE adapter, while this is not going to provide the best performance, it is a realistic configuration for a user who does not have a SATA system or cannot separate the disks due to some issue with their system. Tests were conducted on this system with the mirror and single disk configurations to see if there was any loss of performance mirroring disks with ZFS.
From the graphs, it looks like ZFS provides much better filesystem performance. Notice how the read speed peaks for small files (well, it plateaus not peaks…) where it appears to be almost double the speed of ext3. As the filesize increases, performance drops to ext3 peak levels and it maintains that performance throughout.
I think that is significant as predictable filesystem performance allows you to have a better estimate of how your server/workstation will perform.
I wished that I had more time and more hardware to test with, the iozone graphs definitely show some interesting patterns and it would be cool to see if they are limited by the hardware the test was initially run on, or it is a characteristic of the performance of the system.
I also can’t locate Fibre Channel drivers for Emulex HBA I have in workstation. With Solaris “proper” it just worked all the time.
I didn’t even think to look for the FC fabric packages, which are an integral part of Solaris 10. It is not a surprising move considering the target audience (for now).
It likely isn’t redistributable. Sun was unable to secure agreements for many vendors that allow free, unlimited, redistribution of certain components.
One of the key goals of the new OpenSolaris distribution was that anyone could redistribute it in any way.
So what happens to users of that hardware when OpenSolaris gets released as Solaris 11? Will users that followed the HCL, as Mr. Escue suggested, be left out in the cold or will binary drivers get bundled into Solaris 11?
More than likely OpenSolaris will not get FC drivers simply because it is targeted at desktop as opposed to Enterprise users. Solaris 11 will most definitely have them, along with SCSI and iSCSI support. But this is all speculation at this point because Sun could change direction tomorrow.
If Open Solaris is targeted at the desktop, and not developers like SXDE and enterprise servers like every other version of Solaris, let me predict the outcome:
FAIL.
You seriously can’t think that between Ubuntu, Blue Curve/RHEL, Gentoo, Windows and OS X, this thing has even a remote shot in hell?
Note all 3 work on (FC, SCSI, iSCSI) on RHEL.
This divided strategy is a proven failure by an OS with a lot more users: Fedora/CENTOS/RHEL.
Redhat tried to divide RHEL into WS/Server/etc.
Nearly everyone uses CENTOS, since it does not divide.
So what happens to users of that hardware when OpenSolaris gets released as Solaris 11? Will users that followed the HCL, as Mr. Escue suggested, be left out in the cold or will binary drivers get bundled into Solaris 11? [/q]
Sun will be providing alternate network repositories that allow access to non-redistributable resources.
In addition, the long-term support version of OpenSolaris will likely have additional options that are not currently available.
However, only Sun knows what they will do for certain.
I would encourage you to post about this issue on opensolaris-discuss.
Trust me. Just go look up the pathetic ridiculous fighting over megaraid_sas and mfi.
They have TWO choices, plus they could simply rip the stuff right out of OpenBSD, NetBSD or FreeBSD, but someone actually did the heavy lifting and ported it for them. (Some random dude took mfi from openbsd and ported it lickety split).
Yet it took until svn_89 or something to get in. This has been going on for at least 18 more like 24 months. Check the Sun forums for a huge amounting of whining over PERC support.
Now try and get javaws for the Linux 64 JDK/JRE or the browser plugin for Linux 64. Its been 3 years (Since 1.5.0 FCS) since people started whining about that.
Sun has a lot of bureaucrats really choking the place. Lawyers and idiot PMs. Generally the engineering is top notch, but the branding and direction goes from rudderless to so so.
This is the key issue I see with Sun nowadays.
Do you really need to know anything else?
Because a filesystem on its own is not enough for people to start praising OpenSolaris. There seems to be a misplaced automatic assumption that the mere mention of ZFS will be enough for everyone and cause mass drooling.
Well, most do, and the Unix-like operating system world has moved on now and better usability isn’t just about being root. This is long overdue for Solaris, and it isn’t really news.
That’s as clear a case of networking not working as I’ve read. I’d be very disappointed if I had to resort to that on a Linux distribution (and I’d expect to see things like supplied hostnames handled OK as well), and it’s a big reason why people use DHCP in the first place. The fact is, Ryan Paul couldn’t get any of the networking tools to actually work and give him network access.
Getting up-to-date software in the Linux world is a perennial problem as well. Making this problem worse isn’t going to help OpenSolaris.
I hear that argument all the time, but in practice it tends not to hold up. Software moves on, bugs and security problems are fixed in successive versions and general quality gets better. There’s no evidence to suggest that using Gnome 2.20 is better than using Gnome 2.22 unless you’re willing to backport fixes, as all the code goes into the new version. It still has all the same bugs and problems as it ever had, and having a very long incubation period doesn’t solve the problem. Debian has this problem as well, as software versions they use become unsupported.
As long as you have a reasonable incubation, integration and testing period for your system, using a two year old piece software that no one is committing code to won’t help you make things more stable.
I think it’s fair to say that Sun’s own OpenOffice developers don’t use OpenSolaris as their development platform, and it shows.
The problem is that Solaris usage has declined dramatically over the past ten years, while Linux usage has soared, mainly because of a burgeoning enthusiast community. It’s relatively easy to get Linux into the hands of someone, and demand has tended to shape how your average Linux distro looks today – the bash shell, NetworkManager, HAL, graphical installers etc. Many have just gone out and coded what they felt was needed.
This is why Sun have been pretty much forced into creating OpenSolaris. Leaving it as it is isn’t an option if Sun wants it to be a viable financial investment unfortunately. It’s not a great thing to say, but that’s the bottom line.
It doesn’t ship with OpenOffice because it has to fit on a single CD and there isn’t any room for it right now.
It doesn’t ship with the latest and great applications because OpenSolaris is FCS all the time. Meaning, the goal of the engineers is to be production ready all the time. Sun, unlike most GNU/Linux distributors doesn’t simply throw a bunch of packages into a distribution and “call it good.” Engineering and documentation evaluation is done for every single component shipped and that means things move somewhat slower.
Then, it should be fine, as Sun does backport many fixes and adds many fixes of their own. Sun doesn’t just ship the raw bits that someone else produces.
Wrong. Many of Sun’s developers do use OpenSolaris, and they know they can easily install OpenOffice just by doing:
pkg install openoffice
While Sun definitely has room to improve things, things are nowhere near as dire as you would like to paint them.
Wow.
If you’d read what I’d written, using software that is several versions behind the current version does not make you ‘production ready’ and does not give you any less bugs unless you’re willing to backport vigorously as Debian do. You also end up having a diverged codebase that produces bugs not seen upstream. ‘Production ready’ means absolute jack and is generally just used as cover for this development model.
Well, they obviously have with this release because networking is a bit of a fundamental ;-). I had the same issue, and I was slightly astonished at what I had to do. I’ve never seen a system not just get an IP address from DHCP when asked. If that was a non-production ready Linux distro being reviewed it would be given a big thumbs down all round. I don’t know what the article adds to that fact.
I’ll come to you when I want to find some information on Sun’s site then ;-).
Then what happens there is that you’ve effectively forked the software, as Debian does. This means that you’re responsible for maintaining the software yourself, and in a project such as OpenSolaris that is trying to get more contributors and use more open source software and share resources for its own sake, that just seems a bit…….daft. At this point in time it isn’t what Solaris needs.
Sorry, but that’s just wrong.
And yes, they do backport vigorously.
Networking is supported on thousands of configurations. Unfortunately, in the PC world, there are millions of configurations.
On both my Desktop and Laptop, networking works just fine.
DHCP works just fine as far as I know. You haven’t filed any bugs that I’ve seen, so I would suggest that you do so that any issues can be resolved.
Look at docs.sun.com, etc.
No, that’s what happens when you promise your customers that you will support them year after year. Enterprise-level distributions provide a certain level of stability and support.
In the future, you’ll see OpenSolaris move to having six-month “bleeding edge” releases and a separate long term release to better streamline things.
Until then, complaining about software that is not that old (from a release perspective) is counter-productive.
As for the issues you’ve encountered, unless you file bugs, your complaints aren’t useful.
Edited 2008-06-04 14:44 UTC
Could you care to explain this?
“using software that is several versions behind the current version does not make you ‘production ready’ and does not give you any less bugs”
I thought it is a bad thing to run the latest bleeding edge software on production systems? But, hey, I am no sysadmin. What do I know?
When doing things right, you stage and test things before relying on them. Solaris zones are a big plus, along with staging hardware and a good suite of tests to help vet the software and hardware in action.
Just to know how I feel as a long time Solaris user – OpenSolaris 2008.05 is not only not production-able, I doubt this lineage of software ever will be.
My initial thought is you haven’t been around very long then if you haven’t had issues with DHCP clients. I don’t know you, so I won’t actually say that, but I’ve had to fight with DHCP on Linux, FreeBSD, IRIX, and Solaris in the past. Granted, Linux and FreeBSD are a lot cleaner with DHCP nowadays, but there used to be some serious issues.
On the flip side, I’ve had all kind of trouble using static IPs (!!) with Fedora 9. NetworkManager seems to get rather confused, my network adapter (Intel built-in on a Dell Latitude) doesn’t always show up in the GUI, and I had to tweak a few things to have the interface come up on boot. I need to look for a bug report on that one…
No it’s not.
But it isn’t.
This has never been true. Let’s go back to day one. Sun never even implemented SysV packaging properly, for example. Sun’s own docs on BigAdmin make reference to the fact that Sun sed is horribly broken. Yet it is broken to this day, and they won’t bundle gsed with the OS.
JumpStart and flasharchives as you know them today are not coming back from what I understand. Similar functionality will be provided, but the previous tools, are not equipped to deal with ZFS root, etc. properly. I have also been told that parts of these tools are “encumbered” by third-party licenses.
But fear not, there will be an equivalent, eventually.
I would encourage you to install Visual Panels from the pkg.opensolaris.org repository. Visual Panels will be replacing SMC.
There is no space left on the CD for an office suite. As it is, there is just barely enough room on the CD for what is there. Unlike many GNU/Linux distributions, Sun doesn’t split their 32-bit and 64-bit platforms, and they have better multi-lingual support on a Single CD.
Since many users have downloaded and installed OpenOffice from pkg.opensolaris.org (such as myself) I’d like to know how and what you did to try to install this.
Please report the specific steps to reproduce your desktop issue by filing a bug at http://defect.opensolaris.org/
Cheers.
Thanks for the info on the replacement for JumpStart and Flash, I live and die by JumpStart and Flash. I also was not aware of Visual Panels, I’ll have to check that out. And finally, my bad for not submitting a bug on the OpenOffice issue.
Most other OSes have moved to DVDs by now, or at least offer the option. We saw at JavaOne what happens when you count on always having a fast net connection available.
Why not? It seems like a sensible idea.
A startlingly large majority of Sun’s customers and potential customers surveyed apparently have CD-ROM drives instead of DVD drives. In fact, many of them don’t have either!
In addition, since one of the target audiences was developing countries, keeping the install medium minimal was important.
I suspect Sun may end up offering an alternate version that provides the full content of the CD with an on-disk repository eventually.
However, only they know what their plans are long-term.
Why not? It seems like a sensible idea. [/q]
Because eventually, it is 32-bit support that will no longer be needed.
And, in my opinion, because it makes all platforms first class citizens.
Unlike most GNU/Linux distributions, Solaris supports a mixed 32-bit/64-bit environment that is virtually seamless to the user.
I was asked by Sun about the DVD only option some time ago and I responded that Sun needed to consider the customers who use older hardware that either does not support a DVD drive, or that the necessary drive no longer exists. A lot of people within DoD tend to hang on to their older hardware for years. I am still seeing Ultra 1’s, 2’s, 5’s and 10’s slowly being sent out to DRMO by some and others scarfing them up because the projects they work on have no budget. Of course Sun went with the DVD option, but that is no big deal if you download the ISO and use lofiadm or use JumpStart.
For people who don’t have a lot of experience with Solaris, they wouldn’t know about being able to run 32 and 64-bit apps with no complaints and no real need for separate versions based on architecture.
Because eventually, it is 32-bit support that will no longer be needed.
Its shocking to me that the company which invented the first implemented 64 bit OS and did it very well, far better than say, WOW64 crud on Windows, that the x86 installer up to solaris 10u5 is still 32-bit in x86 mode.
If sun is not going to offer a 64 bit installer for Solaris 10, but gives out DVDs of Solaris 10, I really dont get how the “old machine” crap is a valid excuse.
And yes, you do need a 64-bit installer unless you want a system with a dedicated boot disk, (which is sane), because the 32-bit solaris 10 kernel gets pissed at partitions larger than 1TB – for no valid reason.
So what? LC_ALL=C. Done.
Oh, don’t forget that Solaris x86 uses the 32-bit installer on a 64-bit machine, and that, for no valid reason at all, prevents the making of filesystems larger than 1TB. (fixed in OS 2008.05, but 2008.05 is broken, but not fixed in solaris 10u5, so sun knows about it and gives the finger to the base).
Oh, and don’t forget that GUID partition tables are not readable by a number of the family of snu-supplied disk utilities. (still busted in 2008.05).
Oh, and don’t forget UFS’s cluster size goes from 8192 @ 0.99TB to 1048576 at 1TB. (And ZFS is still not able to not fail under heavy loads so we are stuck with ufs and 1MB block sizes).
Oh, and /usr/ucb/ps is gone and SMC doesnt and get this, Java Web Console doesnt either.
LANG=C=true
LC_ALL=C=true
I dont get it, why are you always bashing Solaris as soon you see an article? You can not help to spew out gall on Solaris, you can not control yourself? Obviously, Solaris must have hurt you deep sometime. I wonder, what happened? Tell me. Solaris made you loose your first job? You ran a big site using Solaris, and the site crashed making you a laughing stock? Something must have happened. This focus on Solaris is not normal, nor healthy. Solaris is not the root of all evil. I dont get it?
Wow……. An entire post about me because I don’t see where Solaris is heading, don’t see how Sun will make OpenSolaris financially viable, don’t see how it will attract wider open source developers and don’t see how an article painting over the problems a few reviewers had is going to solve anything?
No, but now you mention it, a few years ago I did see the look on a Sun engineer’s face, who came in on not an inexpensive rate, when he saw the excruciatingly expensive UltraSparc IV running Solaris get totally creamed away by a factor of 4 (pystones) by a very dusty 1.4GHz desktop Athlon machine when running a Python and Zope site – after Python had been recompiled in Forte by him :-). Silence…….
[The GNU portion of Solaris’s /usr/sfw/bin directory is still pretty spartan to this day.]
My employer wasn’t amused either, especially after my standard response (and that of the Python developers) to our Solaris enthusiast was to just use Linux and x86 if you were using open source software. The Sun engineer babbled something about GCC not being production ready, SPARC being more solid and production ready than x86 and Linux being a poor Unix clone, or some line like that. He was never seen again.
I suppose I just find it a bit sad that that attitude hasn’t changed, nor has Sun really changed anything around Solaris as a response, many years after that experience and three years after OpenSolaris’s inception. After all that teeth pulling, I just find it daft when Sun says “It’s just like Linux and open source!” when it isn’t, and when it gets reviewed someone writes a response that says “Actually, you’re wrong, I edited these files to get it working and you should have reviewed this cool new thing instead”.
Reverse psychology mate ;-). This post definitely isn’t normal, nor has it been implied that Solaris is ‘evil’ anywhere. What’s been written is what’s been written, nothing more. The Sun engineer believed I had something personal against him and Solaris as well for some reason. Must be something in the water ;-).
Erm Sun sells the latest and greatest AMD and Intel processor based systems with linux, windows or Solaris that are far better engineered than the competitors.
Talk about stupid trolls.
Bashing Sun for only selling SPARC is so stupid its not even funny.
Why does this sound similar to the tale I was told by a RedHat sales droid about an application that ran poorly on a SunFire 6800, but magically ran better on 10 dual CPU x86 boxes running RHEL?
The devil is in the details, so how about providing some details instead of just telling us a story. There had to be a reason for the poor performance and I would hope that you were running some sort of monitoring tools to capture some useful information.
I just don’t buy “the performance sucked until I ran it on this” nonsense.
facepalm
Sour grapes.
I think the author hit all of the major points where things are “a little weird” in the initial release.
One thing that I did think was cool was how it snapshots the root volume immediately after the install, automatically, so you can roll back to the “pristine” state if you want to.
I really like the default theme, much more than, say, Ubuntu (which isn’t ugly either).
The use of IOzone reminds me; I recently had a chance to use the latest version of Sun’s Filebench, on Linux no less. It’s a great tool. Early versions were kind of bleeding edge and hard to get going on anything but Solaris, but that seems fixed.
Nothing like a multithreaded Postmark workload to bring a 100 MB/s sequential array down to 1.8 MB/s effective throughput
Glad to hear you like FileBench. I’ll note that it is now included by default with OpenSolaris at /usr/benchmarks/filebench.
I’m actually in the process of integrating some Linux changes submitted by Richard McDougall. OpenSolaris, Linux, and OSX are our three main platforms right now, and we want them all to be reliable.
For those not familiar, check out:
http://www.solarisinternals.com/wiki/index.php/FileBench
Thanks, glad you liked it.
I am not a system admin at all , but I liked your article too…
There is no comparison between an article written for some people that is dedicated to write articles and an article written for technical people that grew and learned some stuff based on his own experience; the later is richer because it destilates wisdom.
Congrats!
Thanks. Forget that I am a system administrator and I have been doing it for almost 9 years. I just find it amazing that people read reviews that are nothing more than a rehash of the Release Notes and think they are great.
To me a review is where the author actually takes the subject of the review out for a test drive and finds out what works and what doesn’t, and writes what he or she experiences during the review process. But I have been wrong before ….
Yeah I am surprised but it is actually decent.
The part about changing the shell and the role is golden. I’ve read the opensolaris forum and have never seen this tip:
The commands needed were not together, first I used the rolemod command as described here:
http://dlc.sun.com/osol/docs/content/IPS/login.html
Then it is simply a matter of using usermod to change the shell for root. I am actually surprised that Sun didn’t have instructions on how to do this, or maybe they figured nobody would want to change it.
The whole article can be a lot shrorter. It would look like this: “I have hardware that OpenSolaris supports, I know how to use Solaris so I don’t know why others complain”.
Edited 2008-06-04 17:00 UTC
Why it works for me and doesn’t for others is becuase I build systems with hardware that is supported by all operating systems. I can install any flavor of Linux, Solaris and Windows on the same machine becuase I bought hardware that supports all of the operating systems I just mentioned. The exception to this is the laptop which I pointed out in the article, which had hardware I knew was not supported.
Where I have the disconnect with people who complain about hardware support is why they buy hardware that doesn’t work with the OS they want to use and then complain that the OS doesn’t support it. If the user cannot read the HCL for the operating systems they intend to install on their machine and later find out that something doesn’t work is not the fault of the OS, it is the fault of the user.
“Where I have the disconnect with people who complain about hardware support is why they buy hardware that doesn’t work with the OS”
When someone was buying his computer 2 or 3 years ago he didn’t know if he will try Solaris (or BSD, Linux, whatever) or not.
BTW. OpenSolaris doesn’t support my USB mouse which is supposed to work out of the box (and yes, I know how to configure mouse).
The hardware I have used to test OpenSolaris is several years old, both have ATI video cards. The only “recent” piece of hardware is the Intel Pro/1000 NIC in the Pentium IV box. I have been able to do it by purchasing quality components that are supported by multiple operating systems, so yes you can.
In regards to the USB mouse problem, that has been hit and miss. Several people here have mentioned having mouse issues, I haven’t experienced that problem or I would have mentioned it.
I think that’s unfair. Look at it this way, would you rather hear from someone who actually knows what he is talking about, or someone who just says, “Well, it’s not like what I am used to. What’s wrong with those fools at Sun, why can’t they just be like {insert fav os here}” ?
That’s not the point some people try to portray here, it is far easier to attempt to make someone out a jerk rather than admit that the author may actually be right.
I have been using Solaris x86 for 10 years, so I know what hardware works. And why Solaris x86, Solaris Express and OpenSolaris works flawlessly on my hardware is because I bought supported hardware and didn’t try to use things not on the HCL. I guess for some that is a hard pill to swallow.
I have SunOS 4.1.4 original media here, as well as a Sparcstation 2 4/75, and even a Sparcstation 20 with 4xCPU config.
I have an Ultra 80 as well.
You can’t roll back the clock as far as I can.
About 2 years ago, for nostalgia, I fitted a 50 Pin 18GB Seagate (ST318418N , the “original” Barracuda 36-ES2) into the 4/75.
Lets not try an assail people for being noobs because they complain. You might be suprised.
Ignoring segedunum in Solaris threads is better for anyone’s mental health. His posts regarding *nix can simply be summarized in “All hail Linux!”
Agreed!
As a solaris admin, this has been a bad experience.
I’ve been routinely breaking ZFS with SunVTS 6.4 and 7 when combined with other loads. First they were panics, a few files becoming locked such that the OS needed rebooting, and other misc. things.
2008.05, as a person accustomed to Solaris, is simply not an option and never will likely be an option to Solaris 10. This new layout is more of a reason to move over to Linux, not less.
The installer has taken away a lot of options, the X-server didn’t work correctly, had to use the vesa driver, the kernel itself seems to be improved but the userland is simply a terrible mashup.
A lot of bundled applications dont work anymore, a lot of applications that used to be bundled are gone, I dont understand the motivation for most of the changes, Solaris 10u5 is clearly more usable, and 2008.05 is certainly not useful for production.
I hope Sun doubles back and focuses on keep Solaris 10 alive for a long, long time. I would consider solaris 10 primary, this new stuff secondary. Its really untenable.
I also think that Sun needs to get megaraid_sas or mfi into Solaris 10 and 2008.05 (not that it matters to me) as soon as possible.
If it weren’t for the nice looking future zfs has (its a good effort, no doubt), and the file system snapshots (fssnap) and great volume management from disk suite to zfs, I would jump over to CentOS or an other sustained linux today. (I think dumping, snapshotting and LVM are rather not-so-good on Linux).
Sun needs to hunker down and keep Solaris 10 going as a first class project.
Also, Sun, please, for the love of all that is holy, please include javaws and the browser plugin for 64-bit Linux. Please, pretty please with sugar on top. Its been 3+ years now.
On what hardware are you having trouble with ZFS? And why are you running SunVTS on it? If you are trying to stress the filesystem I would use iozone.
The reason why the OpenSolaris installer took away options is because inexperienced users were complaining to Sun about the installation being too difficult and having to make too many decisions.
What bundled applications are you talking about? Specifics are nice.
And why should Sun write a driver for an AMI device? If AMI wants to support Solaris, then maybe AMI needs to talk to Sun about driver support for their hardware. It simply isn’t Sun’s problem.
On what hardware are you having trouble with ZFS? And why are you running SunVTS on it? If you are trying to stress the filesystem I would use iozone.
The reason why the OpenSolaris installer took away options is because inexperienced users were complaining to Sun about the installation being too difficult and having to make too many decisions.
What bundled applications are you talking about? Specifics are nice.
And why should Sun write a driver for an AMI device? If AMI wants to support Solaris, then maybe AMI needs to talk to Sun about driver support for their hardware. It simply isn’t Sun’s problem.
Warning, seems I’ve been accosted by a Sun apologist… (Note, I’ve been using Solaris / SunOS since 4.1.4 (I have original media)).
#1 Thumper is the hardware.
#2 Don’t tell me how to stress my boxes. Solaris doesnt fail the stress test with UFS, but it does with ZFS, so what the hell does that have to do with IOZONE? I know how to break file systems, thank you, I dont need you to tell me how.
#3 Ok, let me come up with a list for you, do you want it in PDF, PS, HTML, let me know. Since I have to convince you…
#4 Yes, it is. You see, that driver is needed by a lot of hardware. A nice person made an mfi driver, and made it open source/BSD license. Sun’s ridiculous bureaucracy can’t get the stuff together to choose either LSI’s existing and working megaraid_sas driver (it doesnt work so well under load). So between Sun and LSI, we have poor support for megaraid_sas. That hurts sun, so your pro-Sun attitude would be to go an try and get them to close on this LONG standing issue, not admonish people for needing the driver.
Oh yeah megaraid sas works on:
FreeBSD, Linux 2.4 with lots of patches, Linux 2.6, RHEL, FreeBSD, Windows 2000,XP,2003,Vista (64 and 32 bit), etc.
#5 If sun actually thinks people who couldn’t install Solaris 10 and before will suddenly use Solaris 2008.05, you are on crack. Sorry. And sun POed the loyal base. I know no long time Solaris admins who are happy with this.
No answer or snarky comment about the improper implementation of the Sun 64 bit JDK/JRE, no way to blame me for that?
Oh, and I have an active support contract and have patched per support to try and fix the ZFS issues. No go.
You have to say to yourself, I buy Sun product, I am not the enemy. I pay Sun money, please, dont assail me, the customer, for simply trying to tell Sun how to stay on target.
Or go take on Linux head on, and pretend you’ll usurp Ubuntu with Solaris 2008.05 and dash the hopes of CentOS with over 10,000,000 live systems going to the yum mirrors.
We’ll see how that worked out, just like taking on Microsoft did.
Right. SunVTS should not crash the OS.
OpenSolaris 2008.05 is not currently meant to replace Solaris 10 in the enterprise.
Eventually when OpenSolaris code is stable enough it will be made Solaris whatever. May be you should reserve judgment until whatever release after Solaris 10 based on OpenSolaris is released.
Are you having problems with ZFS on Solaris 10 or OpenSolaris2008.05?
What type of OpenSolaris Support contract do you have? What bugs have you filed for the issues? Can you point to the bugs that your have filed or have uncovered in OpenSolaris2008.05?
Then why is Sun wiping all its Solaris 10 SunRay servers internally and installing OpenSolaris?
It sure didn’t sound like that at CommunityOne.
This may seem clear to you but Sun’s marketing strategy is very unclear to me. Solaris, OpenSolaris (never mind stealing that brand name from the foundation), Nevada, Indiana, what? Tell us what the next release is and we can use it.
As it stands Solaris 10 is languishing and no one is fixing it. Instead they are working on Longhorn/DukeNukemForever/Daikatana/Solaris11.
Look at everyone else, they were smart. Microsoft released XP, then a service pack. Vista, then a service pack. Apple releases OS X, then a service pack. Sun, they release Solaris 10, which changes EVERYTHING, and then anemic patches here and there (routinely causing regressions, who QAs this stuff?) and instead of cleaning up the OS (which has some nice ideas) they are off playing with the new shiny thing. Real users are left to fend for themselves.
Or install Linux/Mac.
Companies usually like to test stuff internally before release. I am sure Microsoft rolled out Vista and Apple deployed Leopard internally too.
They do that to get wider testing.
It sure didn’t sound like that at CommunityOne.
This may seem clear to you but Sun’s marketing strategy is very unclear to me. Solaris, OpenSolaris (never mind stealing that brand name from the foundation), Nevada, Indiana, what? Tell us what the next release is and we can use it. [/q]
Yes it is clear you don’t get Sun marketing. Solaris 10 was being developed for years while Solaris 9 was in the market. Sun does that all the time. Nevada is being developed while Solaris 10 is the shipping release. Fixes get backported to the current release. Many Nevada features have been backported to Solaris 10 so have bug fixes.
Sun decided to open source Nevada and call it OpenSolaris. When Nevada/OpenSolaris is release ready it will get a Solaris name. Something like what redhat does with Fedora and SUSE with OpenSUSE.
People run RHEL in production which comes with support just like people run Solaris 10. Fedora gets the latest and greatest and is considered unstable.
Really! Solaris 10 updates are constantly released. Solaris 10 Update 5 being the latest. Which means 5 updates have been released.
Can you provides a list of high priority of bugs that haven’t been fixed?
Solaris 10 has had 5 updates so far (service packs in Microsoft parlance). Mac OS X Leopard has had 3 till date and it wasn’t released that long ago.
When Sun ships systems with OpenSolaris or some thing based on it then and only then is it considered release ready.
Right now Sun boxes ship with Solaris 10.
Edited 2008-06-05 19:47 UTC
I’ll accept that’s my failure. I don’t care. Sun needs to win me over. They can’t prescribe homework and expect me to say oh clearly Linux sucks. Why should I bother trying to figure out what the brand name is that I want?
This is not a new problem with confusing branding/versioning by Sun. Java. JavaScript. Java Desktop System?! Java 1.0…1.1…2…6…7?!
Normal software has stable/unstable/development. You would think Ian Murdoch would have remembered that from Debian.
Wait, so are updates good or bad? They are good when Solaris has them and bad when Apple has them. Normal Sun logic.
How about this. XP, MS managed to fit most of a userland on top of NT. OS X, Apple built an entire f’in OS But Solaris after 5 service packs can’t have a nice GUI network configurator.
OpenSolaris will have one years from now, but not a normal one. It will sneak around behind your back and do magic. Guess if you are a normal Unix admin and don’t need some overthought abstracted Sun invention but just want to be able to select what interface to use or whatever, you are out of luck.
What about design stuff? Let’s see, stupid example, Apple fit a whole new scheduler into 10.1. Solaris still has CPU affinity problems.
Broken by design stuff, like RBAC. Why can’t you say “everyone in this group has this profile”? Why can’t you specify which profile you want to pfexec a command with, if 2 profiles want to use a same command with disjoint privileges? FAIL. This stuff could have been fixed in the first update, but now the opportunity is gone and we are stuck with legacy.
Hey how come you can’t mount a SMB/CIFS share on Solaris?
So then why do you and Sun laugh when people have problems with Solaris 10?
Huh?
Sun didn’t develop Javascript. Netscape did. Shows how much your really know.
WTF are you on about? You said Sun isn’t updating Solaris 10. I pointed out they are just like everyone else you mentioned in your post. Grow up!
How about this? Solaris 10 is not meant for home users it is for enterprises. OpenSolaris configures wireless devices quite nicely.
Eh! OpenSolaris 2008.05 is Sun’s first ever release. You act as if Sun molested you in your childhood.
Most people seem to have no problem with network settings on OpenSolaris.
What? Care to provide and example. Solaris has had a far suprerior scheduler to linux or any OS for ages. Ever since the linux guys were dreaming about O(1) scheduling Solaris has had one. Solaris also has modular scheduling classes from interactive to real time.
Your inablity to comprehend and use a particular technology doesn’t constitute poor design.
Yes you can. Samba has been included with Solaris since Solaris 9.
So then why do you and Sun laugh when people have problems with Solaris 10? [/q]
Nobody is laughing. Its all in your head, pal.
No, you cant mount a SMB or CIFS share with Solaris 9 or 10.
uname -a
SunOS solten 5.10 Generic_127128-11 i86pc i386 i86pc
which smbmount
no smbmount in /sbin /bin /usr/sbin /usr/bin /usr/openwin/bin /usr/dt/bin /usr/ccs/bin /opt/SUNWspro/bin/ /opt/SUNWspro/sfw/bin /usr/xpg6/bin /usr/xpg4/bin /usr/perl5/5.8.4/bin /usr/platform/i86pc/sbin /usr/sfw/sbin /usr/sfw/bin /opt/sfw/bin /usr/X/bin /usr/ucb /usr/java/bin /opt/SUNWvts/bin /usr/local/sbin /usr/local/bin /usr/sunvts/bin
# mount -F smbfs //user:[email protected]/share /mnt
mount: Operation not applicable to FSType smbfs
# mount -F smb //user:[email protected]/c$ /mnt
mount: Operation not applicable to FSType smb
What OS cluster did you install?
http://docs.sun.com/app/docs/doc/820-2575/gdbbd?l=en&a=view
http://sunsolve.sun.com/handbook_pub/validateUser.do?target=Softwar…
Here is all the software groups the package is installed under:
http://sunsolve.sun.com/handbook_pub/validateUser.do?target=Softwar…
Solaris 9 ships with all the packages on the CD.
Are you done trolling?
Edited 2008-06-06 09:16 UTC
This machine, Solaris 10u5, was installed with every single package.
Are you claiming that mount -F smbfs / smbmount works on Solaris 9, I’d like you to state that it does.
I could do a pkginfo or a showrev -p, but that would be a bit lengthy, particularly since everything is installed.
BTW
# pkginfo | grep -i smb
system SUNWsmbac samba – A Windows SMB/CIFS fileserver for UNIX (client)
system SUNWsmbar samba – A Windows SMB/CIFS fileserver for UNIX (Root)
system SUNWsmbau samba – A Windows SMB/CIFS fileserver for UNIX (Usr)
So much for trolling. You dont even understand what Im trying to do.
No you are right. My mistake. Its late.
Smbfs is work in progress for OpenSolaris.
The foul language, frothing and lack of logic is fun to read in review. Its good to see a debate with two people keeping a cool demeanor. The ad hominem attacks really enhance your credibility too.
Let me explain more carefully for you. Sun has a branding problem. They either license their trademark out — as in the case of JavaScript — or dilute it themselves for things unrelated to Java. Most of the time this hurts the impression of Java with end users and entry-level developers. Java is routinely dismissed due to bad JavaScript experiences. I’m sure the same applies even more to anyone who’s had the misfortune to use the particularly bad GNOME packaging that is JDS (remember the first release of the Java Media Player, anyone?).
Yes, and business users never need to connect their business laptop to the business network….
If that’s the case, why does Sun bother with old GUIs anyway? All Solaris sysadmins I know use text installer, text commands, shell scripts.
I’m sure you’re right; please tell me from your Solaris sysadmin experience how you’d accomplish these tasks:
1. I want everyone in group “foo” to be able to run stuff from profile “bar”. I don’t want to have to run scripts whenever the group membership changes.
2. I have 2 3rd-party products each of which create their own RBAC profiles. There is an intersection in the commands referenced by each profile, but each profile wants a different set of privileges? How come I can’t (especially from a shell script that shouldn’t have to know about the full set of packages/profiles on the system if it only cares about one) say “-p profile“, analagous to “-u user” in sudo?
Seems to me Sun makes a lot of things orthogonal to its own traditions and normal Unix usage but leaves a lot of legacy cruft around without much way to integrate.
Cool, how?
$ uname -a
SunOS pegasus 5.10 Generic_120011-14 sun4u sparc SUNW,Ultra-60
$ sudo mount -F smbfs //galactica/files /mnt/files
mount: Operation not applicable to FSType smbfs
$ sudo mount -F smb //galactica/files /mnt/files
mount: Operation not applicable to FSType smb
$ sudo mount -F cifs //galactica/files /mnt/files
mount: Operation not applicable to FSType cifs
$ ls -d /usr/lib/fs/*smb*
no matches found: /usr/lib/fs/*smb*
$ which smbclient
/usr/sfw/bin/smbclient
$ which smbmount
smbmount not found
$ which smbmnt
smbmnt not found
Nobody is laughing. Its all in your head, pal. [/q]
True, I’m laughing that Sun and its fans think insulting potential customers is the way to win a userbase.
Arun quoted this guy Ben Rockwood of Cuddletech.
http://cuddletech.com/blog/pivot/entry.php?id=938
And he agrees with you. Sun does have a major branding problem:
I really hate bashing Sun, but I’ve gotta speak out against Sun’s continued moronic branding. Following in the tradition of “N1”, “Java Enterprise System”, and the horrible replacement of the good brand StorEdge with the misused StorageTek brand (applied to everything from long time Sun Arrays to Adaptec controllers), comes xVM.
Edited 2008-06-06 13:44 UTC
I never quoted that guy. Robert posted the link to cuddle tech.
Seriously man go get some sleep and some perspective.
Again, a non answer. Sun does have branding issues. Always sidestepping. You into politics or haven’t figured out where people who are non technical who can twist and bend things so much are supposed to work?
Thanks, forgot about that one.
So the Java name is applied to programming language/VM, entirely different programming language, GNOME fork, Web/mail server, toaster, floor wax, …
Don’t forget Java Enterprise Edition and Java Enterprise system are something completely different.
Java went 1.0 to 1.1 to 1.2 which suddenly had to be called “2” but then 1.3 came out and couldn’t be called “3” it was “Java 2 1.3”. Now there are 1.6 (2 1.6) and 1.7 (2 1.7), will Mustang be called 7 or 1.7?
Tons of products called “Java” which have nothing to do with it. Different brand names for different release levels of the same product (one of which brand names was stolen from the foundation that Sun spun off for “independence”, I found out; sound familiar, hmm, Blackdown?).
Explain to me how you are going to install OpenSolaris on a Sun Ray? Have you ever used a Sun Ray? I have one sitting on my desk and I can tell you that Sun would have to make Sun Ray Server software available as part of the OpenSolaris build in order to use Sun Ray devices.
When it comes to Solaris 10, as one the 30 external Beta Testers for what was called Software Express in 2003, I wrote an article for OSNews on its development:
http://osnews.com/story/5485/Review_of_Solaris_Express
My review of the “finished product” 3/05 Release:
http://osnews.com/story/9865/Review_of_Solaris_10
The development of Solaris 10 was going along at the same time Sun was making enhancements to Solaris 9, up until the final release of Solaris 9, I believe 9/05. Each release of Solaris not only patches issues from the previous release but adds new functionality as well.
I also Beta Test every release of Solaris 10, and I can see first hand the effort Sun puts into each Release. To say that Sun is letting Solaris 10 “languish” is simply wrong.
As someone who has over 50 servers running various Releases of Solaris, my problem is deciding which Release to standardize on based on fuctionality.
From the developer I know at Sun, sounds to me like you are not accurately representing the situation with back porting to Solaris 10. I know that internally Solaris 10 is being shafted where possible, and the general attitude is “fix it in Nevada” and big customers or severe bugs generally cause a back port of fixes, but I don’t think at this point its a matter of course.
As a guy who is using many many Solaris 10 u3, u4 and u5 units, I do get very worried that things are slipping when it comes to Solaris 10, and I’ve been backing out more and more patches that I need more often.
The experience I have had with Solaris 10 is a far cray from what I grew accustomed to in Solaris 9 and before, so I do feel there is a bit of product mismanagement.
I’ve used various models ranging from the weird oblong bricky-shaped ones that required their own LAN(?!) to the modern sleek ones with USB ports and LCD screens. I’ve also used WYSE and other RDP thin clients, VT and 3270 terminals, PXE/NFS net stations, …
I wouldn’t think a Sun advocate would be so surprised that someone might have actually encountered one of their products in the real world.
I guess Sun has the Sun Ray packages and can install them on their own systems if they feel like it. I don’t know the details; it’s only the rumor I’ve heard but if true it’s telling.
Regarding internal beta testing / dog food, that may well be, but the point remains that users in the real world will be left further in the cold as no one at Sun will have experience with the problems of Solaris 10 in the real world. All they will see is OpenSolaris without even some lawyer or accountant to run into real life problems. OpenSolaris may be great but customers are not going to run it because as you pointed out so well it’s not supported.
Solaris 10 made a lot of sweeping changes from Solaris 9. A lot of these were good ideas but real world experiences were not incorporated into rapid design improvements and now can’t be due to compatibility. Furthermore Solaris 10 was a half job. A lot of stuff was left undone and everything has been put on hold for years now while OpenSolaris gets hacked on. Solaris 11 may be great but it won’t matter if all the users have migrated away in the meantime, especially if Solaris 11 breaks Solaris 8/9/10 compatibility.
Now why is that. I can roll out the latest OpenSUSE across my userbase and they gain a functionality superset. I’ve also done Fedora, CentOS, Gentoo upgrades, same thing. Sun is always touting ABI/API compatibility of Solaris over other Unix/Unix-like OSes. Solaris even has containers! Why don’t you just have the latest update of Solaris 10 everywhere?
Ok, if you were at JavaOne, you wouldn’t know that Sun didn’t want people to use 2008.05. They declared this thing supported and ready to go. The insinuation was production-able.
“Are you having problems with ZFS on Solaris 10 or OpenSolaris2008.05? ”
The problems with ZFS are on Solaris 10 u4, Solaris 10 u5, and Solaris 10 u5 + relevant patches.
The SunVTS should always be used in consort with directed tests as overall system load is more realistic than just bashing the file system.
“OpenSolaris 2008.05 is not currently meant to replace Solaris 10 in the enterprise. ”
Lets not talk about it then. Its beta being masqueraded as something else. For the record, the previous Sun SXDE 1/08 was way, way better than 2008.05.
“What type of OpenSolaris Support contract do you have?”
I more or less have God level access to sun. I know several engineers there and I have the software support contract. My engineer-friends are having a bear of a time filing stuff with bugster outside of their “area”, and when bugs that I find are filed, the responsible engineers often ask for more info or don’t have access to test beds that would let them recreate the issues.
A recent patch to sshd (126134-03) blatantly broke X-forwarding. I’ve been trying like hell to get it withdrawn. Sun is very, very hard to manipulate for the better.
Again, Im a “sun lover,” but Ill never be an apologist.
I wasn’t.
Then why are you specifically complaining about OpenSolaris2008.05.
Sure. Stuff goes into builds that cause regressions. It happens to any OS undergoing active development.
http://www.sun.com/software/solaris/index.jsp
The page makes it abundantly clear that Solaris 10 is for Solaris administrators/ enterprise developers.
OpenSolaris is for Students/Web2.0 and OpenSource developers.
Some bugs are hard to reproduce. If they were easy to do so internally they would be fixed.
What bugs are you having in particular? Some details would be nice.
I find it hard to believe that just running SunVTS on a thumper causes issues on all those releases and it isn’t a top priority bug.
What was the bug id of the bug you filed?
You don’t have be an apologist but simply ranting on OSNews doesn’t give anyone the impression that you are a Sun lover.
Edited 2008-06-05 20:38 UTC
Wow, so which is it? Can’t complain about Solaris 10, it’s too old and stable. Can’t complain about OpenSolaris, it’s too new and unstable.
You don’t have be an apologist but simply ranting on OSNews doesn’t give anyone the impression that you are a Sun lover. [/q]
Clearly he is a wrecker and a counter-revolutionary.
The problem is all your ranting is so muddled it is hard to make heads or tails of what your complaint really is.
I have asked you umpteem number of times to post the bugids you have filed with your buddies at Sun using your “God” Level support contract and you haven’t posted them.
I am going to have to assume all you want to do is troll and not discuss anything constructive.
If you work for Sun, you don’t want me dropping names in here – trust me.
If you do not work for sun, you wont be able to see my bugIDs (if you haven’t noticed, a good number of them are marked unreadable – particularly the embarrassing ones).
Sigh! Typical. If you don’t have bugids then that’s fine. It doesn’t matter if I can view them or not I would like to know.
I don’t need to the names of your buddies.
Are you the reason Anish Gupta quit?
I never claimed any of that, don’t know what you’re talking about.
I didn’t file bugs, not my job to be Sun QA. Tried OpenSolaris, wasn’t impressed, resumed using openSUSE and Solaris 9/10.
Dump OpenSUSE, CentOS is the least broken current free Enterprise Linux.
What is with you people? Why is it always a political issue what I choose to run? How about this, you want me to run your OS, make it so I like it. No, I won’t file bugs or defects or whatever, I already have something that works and I like it.
True. You responded to my response to some one else.
If you weren’t impressed thats fine. But trolling on an OSnews thread speaks volumes about your deep seeded need to bash sun.
Not bashing Sun. Trying to get rid of the apologists that are rotting Sun with cancerous jingoism from the inside.
I like the feature where you can have the console go both the the keyboard and to the serial port at the same time. Its so convenient for debugging. 2008.05 didn’t fix that one. Is that hard to do or something, you know, simply print kernel console output to both places?
Oh, and I’d also like to point out that while X ran on Solaris 10 u5 on the same hardware, I had to write a customer /etc/X11/xorg.conf file for Solaris 2008.05, it would screw up and killing X would result in an extremely persistent restarting of X which lead to the need to kill it over and over again and over again.
I finally fixed the xorg.conf, switch to the vesa driver, and all was fixed.
Worked in Solaris 10u3,4,5 though.
Seems whatever regressions are done on Solaris 10uX are not done for Opensolaris.
Aren’t we a little testy! I’m not trying to tell you how to do anything, I am just providing an option so back off! I have never used SunVTS to stress a disk system.
I have read about people having problems with SATA systems and ZFS, including Thumpers. Does the system fail only during the VTS testing, this is why I mentioned using iozone is so that you can simulate reads and writes that scale upward and see at what point it fails.
And no you don’t have to convince me, I read lots of “it doesn’t have what I want” threads without specifying what it is that’s missing. So what is missing, is it really all that hard to mention a couple of examples?
My bad on the AMI MegaRaid SAS controller, Sun should already have a driver for it because it is listed as part of the equipment installed in the SunFire X4440 as referenced by Ben Rockwood of CuddleTech:
http://cuddletech.com/blog/
The driver should be part of Solaris 10 Update 4 or higher.
“The driver should be part of Solaris 10 Update 4 or higher.”
Its not.
You use the driver from LSI.
http://www.lsi.com/storage_home/products_home/internal_raid/megarai…
http://wikis.sun.com/display/BigAdmin/Solaris+10+ITU+Install+and+LS…
I’ve done this before and done my tests. The megaraid_sas driver did not survive testing, and later on the test system had issues with updates later provided by Sun. Too bad Sun, with all of its glorious technology partnerships and previously included LSI drivers, cant be bothered to simply add a driver for ubiquitous hardware.
What I like the most about Opensolaris’s big mistake is this Ian Murdoch character. Sun hired him to be a hero, to turn things around. I reality, he was a lucky guy, and him and his wife, Deborah (get it Deb-Ian), some lucky bloke who got the right people around him to assemble an enduring Linux system.
What Sun isn’t getting its the “right place, right time and right people” factor. Ian by himself, judging on what is going on at sun, is a deeply flawed individual leading Sun down the road to perdition.
I think Sun has a good chance of hearing the main users of Solaris’ complaints, and despite the best efforts of the jingoist yes men, maybe the better minds will dig their heels in and say, enough is enough!
# uname -a
SunOS solten 5.10 Generic_127128-11 i86pc i386 i86pc
# cat /etc/release
Solaris 10 5/08 s10x_u5wos_10 X86
Copyright 2008 Sun Microsystems, Inc. All Rights Reserved.
Use is subject to license terms.
Assembled 24 March 2008
# find /kernel | ggrep -ie mfi -ie sas -ie mpt -ie lsi -ie aac
/kernel/drv/amd64/mpt
/kernel/drv/amd64/aac
/kernel/drv/amd64/lsimega
/kernel/drv/mpt
/kernel/drv/mpt.conf
/kernel/drv/aac
/kernel/drv/aac.conf
/kernel/drv/lsimega
/kernel/drv/lsimega.conf
/kernel/kmdb/amd64/mpt
/kernel/kmdb/mpt
No, mpt and lsimega do not work with the controllers in question.
So?
Review the thread. I don’t see megaraid_sas or mfi in there.
That driver, supported in literally every other OS out there (Open,Free,NetBSD,Linux 2.4,Linux 2.6,Windows-Everything-32-and-64bit,etc) is needed by a huge number of machines.
Lsi has a driver for Solaris.
Yes, megaraid_sas, version 1.17. I also have a pre-release version of 1.18.
Its available as a DU.
Easily broken by the same load tests that break ZFS, and megaraid_sas, but not UFS on other controllers.
megaraid_sas, mfi or both should be in there, and as you can see from Ben’s page, its not even in 2008.05, never mind u4 or u5. I don’t think well see it in u6 either.
Please note that Sun is supposedly partnered with LSI (as you can clearly see the long past with lsimega and mpt), as well as Dell whose PERC subsystems are largely now megaraid_sas, as well as Intel. (Ive seen support from Sun on Intel Starlake platforms be a little on the spotty side.)
Sun is giving the OS away, so the more potential places the OS runs, the more revenue opportunities they have. Its that’s simple.
These people who buck and kick and apologize for this should be fired immediately, its this deep ineffective bureaucracy that is marginalizing Sun.
This is pretty pathetic.
Linux and the BSDs can come up with quality clean-room drivers for all sorts of hardware, with no support from hardware vendors and often active opposition. They’re built into the OS and supported on a first-class basis as the OS is developed.
How come Sun, who always makes fun of the development quality of these kids in their basement, can’t match it with their engineers, even when they have special contracts with hardware vendors?
Example: Marvell. They make a shit product but their chips are used on everything. Sun and Marvell have had special contractual relationships for years, but for ages Marvell networking cards (even wired!) never worked on Solaris without ndiswrapper and even when they were shoehorned in, the connection would drop and hang to the point where you couldn’t even unplumb the interface and try to reload the driver.
There are many others.
No one’s impressed by Sun’s mudslinging and name-calling any more. Linux and the BSDs have long since passed Solaris in stability, driver support, performance, etc. Sun has all these awesome engineers and all this money and all these contracts, or that’s what they tell us when pooping on Linux, but they can’t even support a network card or a RAID controller.
What is pathetic is you joined OSnews today just to troll. Could you be segdenum in disguise?
Edited 2008-06-06 08:15 UTC
Ad hominem.
If he is such a troll, what is it that he said that was false/untrue.
Nothing.
What is pathetic is you joined OSnews today just to troll. Could you be segdenum in disguise? [/q]
Cute, ad hominem.
An article interests me enough to deal with yet another double-opt-in so I can comment and I get attacked for it?
I saw OpenSolaris at JavaOne, it looked shiny. I would like to see it succeed. But putting blinders on won’t help. How come the response from Sun and Sun users whenever anyone says “I like it but I had this problem” is “you suck Linux sucks you’re a troll we are smarter than you and everyone”.
I am a long time Solaris user and administrator. I don’t use it as my daily desktop, of course, because that’s insanity. But I always have a server handy and I’ve recommended it to numerous clients before for NFS/application servers, because I knew about API/ABI stability from release to release. Clearly that advantage is going by the wayside.
For the record I’ve never posted on this site under any account before. But what do you care if I’m segdenum, Osama bin Laden, or James effin’ Bond?
The way you say something matters. Your way is counterproductive. Which clearly shows your intention is to bash the product and the company. If you really wanted it to succeed you would provide productive feed back to the OpenSolaris community instead of ranting on an OSNews thread.
What ABI/API breakage have you encountered? Give specifics.
Segdenum is a well known Anti-Sun troll here. It makes a huge difference if you are just him/her masquerading under a different name.
Good plan there, piss off the experienced users, with legacy servers/apps, who are the only market you will ever have EVER.
Hey, remember when Carly was running HP into the ground, and they decided they were going to piss off the engineers who would only use HP calculators, and they took out the equation library and moved all the keys around to appeal to high school students? They sure have that high school calculator market sown up now, you never even see a TI anymore.
OpenSolaris might be the new incarnation of Sun Express Developer Edition, but they seem to have removed everything developer-oriented. When a tiny SLAX live CD contains a functional GCC, why can’t OpenSolaris?
“But,” you say, “it’s easy to download the GCC package.” True enough, except that I need ndiswrapper to use my wireless card to download GCC, and I can’t install ndiswrapper without GCC! SLAX includes ndiswrapper too, why can’t OpenSolaris? In the meantime, Sun’s Ruby Developer site still says the following:
> Get NetBeans. Get GlassFish. Get Cool Stack. Or get them all with Solaris Express, Developer Edition.
It’s a good thing that I’m getting my eyes checked today, because I sure couldn’t see any of that in OpenSolaris…
Before you assume that I’m another Linux fanboy coming to whine here, I’m not. I use NetBSD as my primary OS, and I don’t like Ubuntu. But this was a sad experience.
Edited 2008-06-05 19:12 UTC
You can easily use the ethernet port and download what you need or burn a CD with what you need.
I have had to do that many times when Linux was still growing. I have had to do that on windows too.
Solaris Express developer edition ships on a DVD.
That’s because the usage model is you run pkg and suck it from a repository.
http://pkg.opensolaris.org/status
This is what the OpenSolaris 2008.05 download site claims:
“The OpenSolaris 2008.05 Live CD makes it simple to boot to a fully functional desktop environment, including Firefox and Thunderbird, without the need to install onto your system. After familiarizing yourself with the OpenSolaris environment, you can then choose to install it onto your disk. Once installed, you can connect to the OpenSolaris Package Repository to install additional software at OpenOffice.org”
No where does it claim that it will include everything Sun makes in one CD.
Why? Because you didn’t understand the concept? None of the things you bring up are actual problems.
You claim to use NetBSD but are comparing OpenSolaris to SLAX???!!!
If you want all those developer tools but don’t want to download it all after you install OpenSolaris you can download the Solaris Developer Express DVD.
My point with the SLAX example was that a very small OS contains many valuable tools that aren’t in the much larger OpenSolaris. I have to question Sun’s decision to include Compiz but not GCC. I guess it’s an issue of priorities.
Yes. I think Sun is trying to prioritize getting new people who might be familiar with linux to try cool things linux doesn’t offer.
There is no reason you shouldn’t send feedback to the OpenSolaris community about including GCC in the live CD image.
Just put the stupid thing on a DVD like everyone else.
Wow, you guys are pretty harsh. The article is interesting, thanks, and I look forward to future OpenSolaris releases — looks like Sun is really starting to innovate again — but some of these comments verge on insanity.
I’ve seen better handling of criticism by RMS, JWZ, Theo de Raadt, and Hans Reiser on their worst days. Why not fix the bugs, or WONTFIX them, instead of ranting on message boards?
Because so far there haven’t been any real bugs with specifics discussed. 🙂 I am sure none of them had to put up with people that came there just to bash something with no technical detail.
Edited 2008-06-06 08:57 UTC
Funny, I remember using mdb quite a bit. Is that technical? You want 8GB dump files or something? Stack traces?
Bugids.
BugIDs: 6689906 6642929
http://quadium.net/random/comments/kitten-yarn-love-thread.jpg
Harsh? Harsh?
Look, Solaris 11, Nevada, Indiana, or whatever nom-du-jour they come up will probably never get into Sun’s biggest account: The US Government.
So this abandon Solaris 10 attitude could basically finish Sun off for good, why? Redhat Enterprise Linux is now Common Criteria/EAL4+ certified.
So while Sun is busy tearing Solaris to pieces to “make it better,” its pissing off pretty much every old-school and the biggest customers all at once.
http://www.redhat.com/solutions/government/commoncriteria/
Where did you get this idea? Why do you think OpenSolaris or whatever it will be called won’t meet the US Government’s vendor selection criteria?
Is there a reason why OpenSolaris won’t be EAL4+ certified?
Edited 2008-06-06 10:09 UTC
Oh, but its ready now! You can get support contracts for it! I was at Java One, go hither and get OpenSolaris 2008.05, its ready for production!
We’ll see when common criteria finishes. We’ll see if the US-GOV takes the “new stuff” as a replacement for existing Trusted installs.
Time is ticking. From the start 2008.05 has, it wont make it in time.
Edited 2008-06-06 10:26 UTC
Oh Grow up already! Seriously this is getting tiring. Yeah we will see.
You claimed it as if it was a foregone conclusion.
You seriously are acting as if OpenSolaris 2008.05 killed your pet gold fish. Geez!
Edited 2008-06-06 10:35 UTC
Again, you act as this is meant to attack sun, its meant to help. I was at Java One. Most of my friends have worked there in the past. I live about 5 miles from the main campus. I have years worth of old sun gear … and you just cant understand that friends tell friends the things they need to hear, not what they want to hear.
OpenSolaris 2008.05 is the worst “shipped” product I’ve seen to date. That maybe and the broken cache on the on the UltraSparc II:
http://www.theregister.co.uk/2001/03/07/sun_suffers_ultrasparc_ii_c…
Bright idea to release a processor with parity on SRAM/cache and not ECC:
The solution? Double the amount of memory so half of it can be used to “check” the other half.
You follow that by this:
WTF? Seriously Every cpu manufacturer has had bugs. How about intel releasing the pentium with the F00F bug.
http://en.wikipedia.org/wiki/Pentium_F00F_bug
Double the amount of memory? One of the solutions was a mirrored cache ( not all of main memory). The other solution was a software cache-scrubber.
Ok, Intel and F00F. How much money does Sun make on CPUs? How much does Intel? Sun doesn’t have its own fab anymore and is largely migrating off of SPARC after years of SPARC evangelism. I remember trying to convince people at Sun to look at the Opteron before Bechtolsheim returned. I was mocked, told how great the v65 was, and how Sunray was the future. Now by the time Sun gets around to being “competitive” with AMD64, Intel’s back on top with the Core2 (not as scalable, but they certainly are selling a lot of those).
Also note, the cache mistake Sun made hurt Sun a lot more than F00F. I happen to know the deal which was lost due to that fiasco. Things would be different now had that not happened.
Again, if you haven’t noticed, there is little you can bring up about Sun I don’t already know more about, and would, at the end of the day, I would like to see SUN do very well. The direction things were taking with Jonathan Schwartz looked good for a while, but now its not so good. Sun announced partnerships with Dell and Intel and I can come up with a number of machines were Solaris 10 or 2008.05 don’t run properly due to Sun’s inability to learn from what others have done in the open. (e.g. mfi)
And yes, the cache size was doubled in order to create a mirror. That’s an expensive and foolish way to solve the problem.
Now we could talk about the merits of crossbars, firebus or even hypertransport vs. the Xeon FSB/MCH, but seriously, INTC vs SUNW (oops now JAVA).
Whatever was working before is now being broken.
You know how many DEs left Sun to start SUPER successful companies with a larger market cap than sun today? You know why? People like you who defend brokenness to the end instead of filing a damn bug and fixing stuff.
Edited 2008-06-06 11:05 UTC
What? Where is Sun migrating away from SPARC? Where do you get these silly ideas?
Sun never had a fab for SPARC. They have always used TI. http://www.sun.com/processors/feature/texas.html
Sun acquired cobalt and released an x86 box long before Bechtolsheim returned. You do realize that Sun sells intel core 2 based boxes, right?
For a person who claims to know a lot about Sun, you sure don’t know a lot.
Sure. But why is that relevant now?
See above. There is a lot you don’t know about.
May be if you did a search on the OpenSolaris community forums or google you would have found that there is active work going on for integrating mfi into OpenSolaris:
http://www.opensolaris.org/isearch.jspa;jsessionid=56F2069C0614734F…
Why?
Firebus? There have been UPA, Safari and JBus on Sun cpus I have never heard of firebus.
In hardware or software?
I am not defending brokenness. I am arguing with a person that chooses to rant rather than contribute productively.
People leave companies all the time. A lot of people are leaving Google to do the same. What does that prove? Zilch.
Edited 2008-06-06 11:34 UTC
“What? Where is Sun migrating away from SPARC? Where do you get these silly ideas?
Sun never had a fab for SPARC. They have always used TI. http://www.sun.com/processors/feature/texas.html ”
Yes, an TI clearly isnt making SPARC process much better for Sun because it makes billions on DSPs and nothing on SPARCs.
So there were no ROSS cpus? Hrm. I remember them.
“Sun acquired cobalt and released an x86 box long before Bechtolsheim returned. ”
I said AMD64. Opteron. And Sun destroyed one of the coolest appliances ever. I have a RAq4. I know them well. It still works. Sun EOLed them. Good move.
“Sure. But why is that relevant now? ”
Because as silly as that was, doing to solaris what solaris 2008.05 is doing to solaris makes me think about things like that.
“May be if you did a search on the OpenSolaris community forums or google you would have found that there is active work going on for integrating mfi into OpenSolaris:”
Its been years, give it up.
“Why? ”
Yes, SRAM is SOOOOO CHEAP. Christ.
“Firebus”
Grasshopper: See XVR-4000
http://www.sun.com/desktop/products/graphics/xvr4000/details.html
And it was written: Sun Fireplane interconnect, DMA for data transfer (4.8 GB/sec.)” ..
“In hardware or software? ”
Hardware seems ok. Since I can run previous Solaris OSes and other OSes ok, but 2008.05 chokes on my testbeds, Ill take a gander that would be the half-baked software.
“People leave companies all the time. A lot of people are leaving Google to do the same. What does that prove? Zilch. ”
I said, DEs have left sun to make companies with larger market caps than Sun.
Andy, who left sun, happened to angel Google. And he built lots of companies with values far more than sun.
This is what happens when a bunch of real defensive snarky high browed punks stifle creativity and produce mostly irrelevant products and have problems monetizing technology largely in part due to the critical mass of people like this:
http://www.mgroves.com/images/south_park_smug.jpg
What is wrong with you?
You still haven’t answered where you go the idea that Sun is divesting from SPARC.
Gee. Last I checked Sun didn’t own a controlling stake in ROSS. So your “Sun doesn’t even own a fab anymore” is utter BS.
Yes. I retorted with Sun makes intel core 2 based boxes too.
That’s your prerogative.
Bull shit. There are post there very recently. Oh so since your BS got called I need to give it up. Grow up!
Redoing the CPU to add ECC is cheaper, right? Einstein.
That is the marketing name for Safari and its fireplane not firebus.
http://www.sun.com/processors/manuals/External_Schizo_PRM.pdf
The XVR-4000 replaced a cpu module and sat straight on the cpu bus/interconnect.
Andy didn’t start Google. He wasn’t a DE. He was one of the co-founders. What other company did Andy build that has larger mrket cap than Sun? Name one DE that has started a company that has a larger market cap than Sun?
Nothing.
Largely incoherent ranting by people like you who think they know better. Come to think of it that image describes you perfectly.
“You still haven’t answered where you go the idea that Sun is divesting from SPARC.”
Im not seeing new sparc rigs anywhere I go, and I get into a lot of places.
“Gee. Last I checked Sun didn’t own a controlling stake in ROSS. So your “Sun doesn’t even own a fab anymore” is utter BS. ”
Sorry, Even back in the Moto days, they didnt have a fab. So the BS thing is just you trying like hell to find anything to discredit. Its like watching a fish flopping around inside of a fisherman’s boat.
“Yes. I retorted with Sun makes intel core 2 based boxes too”
A bit late on the stage IMHO.
“That’s your prerogative. ”
No, not just my “prerogative” – I cant find any admins I know that are “getting behind this”
Remember ADMINS run Solaris, not freaking PROGRAMMERS.
“Bull shit. There are post there very recently. Oh so since your BS got called I need to give it up. Grow up”
You are a complete fool and are being rather Assy. Look for PERC SAS and Solaris on Sun forums. This problem has been languishing FOREVER. And comes back if it gets solved temporarily. And Sun is ‘partnered’ with Dell.
“Redoing the CPU to add ECC is cheaper, right? Einstein. ”
Taping out broken crap is what the problem is here. Same attitude that begat the SPARC II begat 2008.05.
“That is the marketing name for Safari and its fireplane not firebus.
http://www.sun.com/processors/manuals/External_Schizo_PRM.pdf
The XVR-4000 replaced a cpu module and sat straight on the cpu bus/interconnect. ”
Sorry, fire plane, firebus. Im not using google and PDFs and sun.com to *try* and look smart, I know this stuff.
“Andy didn’t start Google. He wasn’t a DE. He was one of the co-founders. What other company did Andy build that has larger mrket cap than Sun? Name one DE that has started a company that has a larger market cap than Sun?”
Andy angeled google, fool. He wrote a check to angel google for 100k before google inc. existed.
Mike Malcom (Netapp) was a DE. And Andy, well, he was employee #1. Id say he was a DE before DEs existed. So that company #1: Netapp. Sun is suing them to get “their IP” back. Lol.
“Nothing. Largely incoherent ranting by people like you who think they know better. Come to think of it that image describes you perfectly.”
Seriously man, I haven’t handed someone’s butt to them as bad as I have to you in a while. QED.
Edited 2008-06-06 12:09 UTC
Eh! How does that have any bearing on Sun’s SPARC road map. Sun is actively releasing new processors like the UltraSPARC T1, T2, T2+ and the ones with the Fujitsu collaboration.
Come on be ready to back the ludicrous claims you make.
The problem is you don’t say much stuff that is credible. I asked you for bugids and you don’t provided then. You claim “Sun doesn’t have its own fab anymore and is largely migrating off of SPARC after years of SPARC evangelism. “. But back peddle when your BS is called out.
You are entitled to the,.
I should trust this just based on your word. Like I should trust your claim that you actively filed bugs for all the issues you found because you are Sun’s friend and have “god” level access to support. Riiiggght!
LSi has a driver that is supported on Solaris. Your beef is that mfi isn’t in yet. Yet you make it sound as if Solaris won’t even work on Dell boxes!
Bullshit. It isn’t broken if it wasn’t designed in you idiot.
Dude you couldn’t even if you tried.
You blithering moron. You claimed DEs left and “started” companies.
Angeled is a far cry from starting a company you dimwit.
What utter crap? Sun isn’t Suing Netapp. Get your facts straight. NetApp is suing to block ZFS because they are shit scared. “DEs before DE’s existed” Geez, that’s a stretch.
Netapp isn’t larger in market cap than Sun. Care to make any other wildly idiotic claims?
LOL. Do you always act this way when your ass gets whooped around in circles?
”
Eh! How does that have any bearing on Sun’s SPARC road map. Sun is actively releasing new processors like the UltraSPARC T1, T2, T2+ and the ones with the Fujitsu collaboration. Come on be ready to back the ludicrous claims you make. ”
I remember something about a CPU recently being canceled. Is Rock quad core or 16? Anyways, like I said, im sure those are super high volume and Sun is making loads on that and nothing on support and government sales. (not). Dell shakes in its boots at Sun.
“The problem is you don’t say much stuff that is credible. I asked you for bugids and you don’t provided then. You claim “Sun doesn’t have its own fab anymore and is largely migrating off of SPARC after years of SPARC evangelism. “. But back peddle when your BS is called out.”
Sure, my credibility is shot- riight. I’m withholding the bugIDs, I might post them later to prove you wrong. I’m sure you’ll refresh the page a lot.
Because without those IDs, you’ve been debunked and have not effectively handled anything. You seem to need the bugIDs, I dont. They are filed. I can reproduce.
And whenever you can you do ad hominem because you have nothing. Wouldn’t it be rich If I posted up a Sun badge…
“I should trust this just based on your word. Like I should trust your claim that you actively filed bugs for all the issues you found because you are Sun’s friend and have “god” level access to support. Riiiggght! ”
Like I said, this is orthogonal to most of what is being said here. Ill post them much later, to ice the cake and keep you in limbo. I’m looking at them right now lol.
“LSi has a driver that is supported on Solaris. Your beef is that mfi isn’t in yet. Yet you make it sound as if Solaris won’t even work on Dell boxes!”
As I sad, megaraid_sas, which you didnt even know the name of, fails on testbeds. Me so sorry, you fail at dis argument too.
“Bullshit. It isn’t broken if it wasn’t designed in you idiot. ”
So Sun designed a broken CPU on purpose? Parity hangs on a huge system Smart. Good simulation job. I happen to work with FPGAs at the moment and laugh at this type of mistake. So preventable. and there are like thousands of employees, lol.
“Dude you couldn’t even if you tried. ”
I have the hardware, I have the admin jobs and the money. Sorry that I, a real consumer of Sun product, and quite a bit of it in the days past, is being savaged by a Sun fan boy. Rich.
“You blithering moron. You claimed DEs left and “started” companies.Angeled is a far cry from starting a company you dimwit. ”
Ok moron.
– Andy Angeled Google.
– Mike Malcom started Netapp.
Christ, you are dense.
Oh, I think sun started it. Anyways, its fun to watch a company that makes a lot of money go at it with a old dog trying to learn some new tricks.
Now you DEny Mike Malcom.
Oh yeah, and Netapps MC is 8bn, Sun is actually up to 9BN, impressive, its up from the 2.00 a share days, I forgot, but did you happen to notice that SUN has 35000 employees vs Netapp’s vs NTAP’s 4000.
Its funny – SUN buys storagetek and makes ZFS (incomplete as of yet in terms of production-ability) and it wants to take on NTAP which has been hugely successful for years.
Seriously Arun. If you work for SUN, you are the worst PR ever, and you appear rather unintelligent and lacking background.
If you don’t work for Sun, you have a serious problem – sun wont hire you for acting like a moron on forums.
Edited 2008-06-06 12:54 UTC
I wonder, when can I look to see Sun making *anything* remotely profitable (or affordable) that gets a top score on spec.org ? You think it will be the Ultra-SPARC V? Gemini? Oh that’s gone too!
So get this, the top performer in CINT, well, Sun does well with an X4150, only to be bested by, get this, a:
ProLiant DL380 G5 (3.16 GHz, Intel Xeon X5460)
http://www.spec.org/cpu2006/results/res2008q1/cpu2006-20080218-0342…
http://www.spec.org/cpu2006/results/res2008q1/cpu2006-20080219-0346…
Now for CFP, Sun is again in 3rd with a hussied up Fujitsu rig costing BOATLOADS of money, and in comes a little cheesy Supermicro, just behind.
Supermicro X7DB8+ (Intel Xeon processor X5460, 3.16 GHz)
http://www.spec.org/cpu2006/results/res2008q1/cpu2006-20080304-0360…
http://www.spec.org/cpu2006/results/res2007q4/cpu2006-20071126-0268…
Its fun to watch off the shelf hardware give Sun/Fujitsu such a hard time. But you know who sits out WAYYYYYYYYYY in front, not caring, is IBM. You see, they dont worry about Dell and Supermicro… But Sun sure has to.
“Do you always act this way when getting whoopedin circles”
Do YOU (Since YOU seem to think YOU are JAVA(SUNW) always act this way when YOU are getting YOUR butt whooped?
Seems like NTAP yoy growth is +17%, and Sun’s is -0.50%.
http://finance.yahoo.com/q/ks?s=SUNW
http://finance.yahoo.com/q/ks?s=NTAP
JAVA (SUNW)
Qtrly Revenue Growth (yoy): -0.50%
Stock Price History
52-Week Change3: -36.82%
S&P500 52-Week Change3: -5.81%
NTAP
Qtrly Revenue Growth (yoy): 17.00%
Stock Price History
52-Week Change3: -21.02%
S&P500 52-Week Change3: -5.81%
– Sun Market Cap, 35000 employees, 9 Billion
– NTAP Market Cap, 6800 employees, 8 billion
Look at all the positive press 2008.05 garnered for SUN.
http://www.247wallst.com/2008/06/a-viking-funera.html
June 03, 2008
A Viking Funeral For Sun Microsystems (JAVA)
Sun Microsystems (JAVA) hit a new 52-week low of $12.32. The market has believed for some time that the server company will have a poor year with relatively flat revenue and little or no operating income.
Sun has not been part of the recover in tech stocks. Since the beginning of the year, JAVA is down almost 30%. Shares in HP (HPQ) are off a little over 7% and IBM (IBM) is up 10%.
Wall St. is unlikely to bid the price of the shares up unless the company’s board is prepared to replace management or break the company into pieces. Sun has three core business: hardware, storage, and support and professional services. In the last quarter, product revenue for the company was $2 billion and service sales were $1.26 billion.
Sun’s shares have nowhere to go, if the firm continues to be operated in its current form.
Douglas A.McIntyre
Yeah, JAVA(SUNW) is sure scaring HP, IBM and NTAP.
Edited 2008-06-06 13:11 UTC
Usually the defeated, particularly in battles of intellect, resort to name calling, foul language and ad hominem attacks. Its akin to the id in a 2 year old child taking over and watching a classic tearful tirade.
Its Information Minister Arun Muhammed Saeed al-Sahaf (aka Comical Arun) declaring that Mike Malcom is not ex-Sun and Andy (EID1) Bechtolsheim didn’t Angel Google for 100k before Google Inc existed.
When did I say Mike Malcom is not ex-sun?
Copy Paste the exact quote. Or shut up.
I didn’t realize what I was getting my self into. There have been 30 posts since I last left and 99% of it are from you when no one else is on the thread. Wow you need professional help.
I am done here.
Concurr, this guy is worse than shaman ever was!
Let’s say we stop feeding the troll. He seems to be doing fine with no fodder.
Smarter. Its not a troll though when you get outsmarted.
Done.
Not really. Google is building quite a cache of this.
Done? Big loss, since you were so helpful, such a great representative and advocate for Solaris, and providing such a wealth of information and answers and not being argumentative at all.
Done? Big loss, since you were so helpful, such a great representative and advocate for Solaris, and providing such a wealth of information and answers and not being argumentative at all.
jongoist (n) jingo·ist – Arun.
Attack the man, not the idea. Keep icing the cake.
Not yet. In a few days or so the bugIDs will be here – waiting. You need them as a last shot at drilling any holes in anything that I’ve said… Can you resist? Do you want one last shot at proving yourself “right” when shown to be so wrong? You should re-read what you said, because you were button mashing and frothing, you mis-stated and used vicious profane and ad hominem attacks in responses. And I need help? Good one.
Just for posterity:
The first ad hominem using profanity between Mick and me was started by Mick:
by mickrussom (1) on Fri 6th Jun 2008 12:07 UTC in reply to “RE[10]: wow, you guys are pretty harsh”:
“You are a complete fool and are being rather Assy.”
Edited 2008-06-06 16:14 UTC
Nope. You started getting nasty first with the other guy. Go read in chronological order. And the minute you REQUIRED the use of ad hominem because you had nothing else, I tried to explain things to you reasonably.
Your were rather frothy and in poor form. Let this serve as a record.
Wow! LOL What business is it of yours if I talk a certain way to some one else? That person can take care of themselves.
The record shows that between you and me things got uncivilized after you started the name calling. Nuff said.
The record also shows that you’ve been talking to your self for the past 3 hours.
No, the record shows clearly where you got rude to more than one person.
And I wasn’t talking to myself, I was on the phone with friends and this was a blast to read.
You talk to your friends on the phone by posting 25 odd messages in the ether on OSNews. Wow! simply wow!
No, no wows for me. You are proving to be the most amusing. I’m correct and you are funny.
And still, problems are easily found during stress tests of ZFS….
AND DID YOU NOTICE, During Java one that when they smashed the hard drive with the sledgehammer, they pulled the sata cable out real quick!?!? Thats a HINT at one of the biggies I found in ZFS………….
(If you leave the cable in and drive is not responsive strange things happens every so often!)
http://www.youtube.com/watch?v=CN6iDzesEs0
At 4:50, the conman surreptitiously pulls the cable from the dead drive.
There is a bug logged for this. Half dead drives on the cable cause MAJOR issues, and no, you don’t even have to smash them.
Thats not “THE BUGS” I could show you regarding ZFS and a few other things. But its a biggie.
Wrong again on who is first.
by Arun (2) on Fri 6th Jun 2008 11:56 PST in reply to “RE[9]: wow, you guys are pretty harsh”
Nope:
Gee. Last I checked Sun didn’t own a controlling stake in ROSS. So your “Sun doesn’t even own a fab anymore” is utter BS.
Called something in reply to Ross CPU, “Bullshit. ”
Bull shit. There are post there very recently. Oh so since your BS got called I need to give it up. Grow up!
More bullshit.
Redoing the CPU to add ECC is cheaper, right? Einstein.
Calling me “Einstein” facetiously.
Largely incoherent ranting by people like you who think they know better. Come to think of it that image describes you perfectly.
I could go on.
No matter, when facing an al-Sahaf-like individual.
Ad hominem means attacking a person. It would help if you new what it meant first. Bullshit isn’t a personal attack.
“Bullshit (often abbreviated BS), also Bullcrap, is a common English expletive. It can also be shortened to just “Bull”.
Most commonly, it describes tautological, incorrect, misleading, or false language and statements. ”
http://en.wikipedia.org/wiki/Bullshit
Einstien is analogous ending something with Smart guy and can hardly be considered a personal attack. But that’s your interpretation.
Calling someone a “complete fool” is a personal attack.
well, i think the record shows that the ad hominem started not with me but with jjgorsky and you were so quick to be nasty you apparently were confusing replies. now since you started being nasty, foul and rude first, you are trying to get technical about ad hominem, which means “argument against the man” – now you are trying to draw the line where it favors you. History will show you crossed the lines of decency first. I’d love to hear a Kieth Olbermann rant on you after this whole ordeal, he couldn’t stop ribbing you for how loose you are with the truth!
http://www.youtube.com/watch?v=CN6iDzesEs0
At 4:50, the demo-giver surreptitiously pulls the cable from the dead drive.
There is a bug logged for this. Half dead drives on the cable cause MAJOR issues, and no, you don’t even have to smash them.
Seriously, If you would just slow down and think, you wouldn’t get p0wned.
Say what you know, and don’t speak or back off when you don’t. Stop trying to Google all your retorts, slow down and just try and put yourself in someone else’s shoes. You come off as arrogant and mister know it all, and you come of as a rude neophyte. The old guard has seen this time and time again.
BugIDs: 6689906 6642929
Come to think of it that image describes you perfectly.
That image is of happycat, Morda the cat from a Russian cat food company (Хеппи Кет). I don’t know how pictures of pets “describe” people, but hey, you aren’t making much sense.
That faceless drone non-entity default avatar is good for you! Keep it.
http://www.osnews.com/images/user.jpg
Edited 2008-06-06 12:15 UTC
You comprehension skills are asinine. I mean the south park image you mentioned.
Oh, sorry, you were, as usual, ambiguous. I just assumed you were attacking me personally in some way again.
Well, I didn’t mention it. I presented a link.
“May be if you did a search on the OpenSolaris community forums or google you would have found that there is active work going on for integrating mfi into OpenSolaris:”
I was thinking about this. It took one guy a few weeks to port OpenBSD mfi to Solaris. Its taken sun on the order of years to do anything with either mfi or megraid_sas.
The competence is evident.
Porting a driver and actually making sure it release ready are very different things. Your ignorance is very evident as well.
Edited 2008-06-06 11:58 UTC
No, I appreciate the difference and have been begging to get sun to get it into product//regression for a long long time.
Porting a driver and actually making sure it release ready are very different things. Your ignorance is very evident as well. [/q]
So…. speaking of OpenBSD…
How come Solaris’s SSH is so broken, much more broken than the OpenSSH port to Solaris (build from openssh.com or get from sunfreeware). No ForwardX11Trusted, PermitUserEnvironment is supported now but the –with-default-path is anemic and there’s no proper way to set default environment vars for users (rsync via ssh fails without –with-rsync-path), no ControlMaster, etc.
And yet as mickrussom pointed out before, Sun broke X11 forwarding recently and apparently has yet to withdraw the patch. No one’s bothered to set up an automated regression test for X11 forwarding via SSH yet?
Yeah, Sun sure has something to teach the OpenBSD people about release readiness.
Yes, I’ve been trying like hell to get: 126134-03 withdrawn.
Its completely broken, breaks X-forwarding and in no way can you enable it with tricks, forcing IPv4, you name it, its broken.
Anyways as of 6/6/2008, its still there.
sigh.
Where are the ZFS bugs? You keep brining up the wrong bug.
In due time. Since there is *nothing* that you can debunk, the bugs IDS, I dont really need them. You seem to.
Just go to bugs.opensolaris.org yourself and check the open accepted high priority bugs in ZFS. Dont be such a lame troll, the evidence is there for the taking.
Edited 2008-06-06 12:48 UTC
Oh, and that particular sshd bug is OK with you. Your commitment to quality seems unparalleled.
Yep, it not only has bugs, but under load, the runQ is a lot longer and the %sy vs. %us is higher (sy is a LOT higher) when compared to UFS under the same loads.
So what you get is more time spent coddling the file system, and you’re getting it wrong.
Thanks. Now if Sun wants ZFS to save the company, fix it up. In the meantime, while we still have to use UFS, can Sun please put the UFS over 1TB block size to something less than 1048576. Maybe 16384 or 32768. That would be a good compromise.
Didn’t see it in 2008.05. Now we aren’t even trying it seems. If mfi was kept out for “issues,” why was all the horrible brokenness that got in there allowed in?
“People leave companies all the time”
The reason is important. Right now, there is brain drain at SUN for a few reasons. People like you, and the lack of a real upside for the stock.
You guys are very nice to Uncle Sam:
Sun Cancels GSA Contract After Price Abuse Probe
Washington Post Staff Writer
Saturday, September 15, 2007
Sun Microsystems, one of the government’s top technology vendors, canceled its contract with the General Services Administration after months of questions from investigators and lawmakers over allegations it had overcharged taxpayers in recent years….
Maybe you wont need common criteria after all.
http://www.washingtonpost.com/wp-dyn/content/article/2007/09/14/AR2…
Do you have any idea of what you are talking about when it comes to Common Criteria evaluations? It is unlikey that OpenSolaris would ever be evaluated under Common Criteria in much the same way that Fedora or OpenSuSe has not been evaluated. Sun, RedHat and Novell aren’t willing to spend $500,000.00+ to have an evaluation conducted on what is not intended for DoD consumption.
The release cycle would have to be significantly slowed down to support CC evaluations of OpenSolaris. That is why Sun is developing Solaris 11 along with Solaris Express and OpenSolaris, there is no point in evaluating OpenSolaris.
Solaris 10 11/06 has been evaluated at EAL4+:
http://www.commoncriteriaportal.org/products_OS.html#OS
And it really doesn’t make any difference whether RHEL is EAL4+ or not, one thing I am still waiting for is RedHat to show administrators how to use auditing based on DISA requirements. While they point out that auditing exists, there certainly isn’t a clear set of instructions on how to either (1) use it or (2) a guide that compares and contrasts to BSM auditing for Solaris. All you hear from RedHat is SELinux, which I think is nothing more than snake oil. Remember Windows NT was also evaluated as TCSEC C2 (EAL4 equivalent).
Well, if Sun keeps trying to cheat the government, the common criteria may not matter.
http://www.washingtonpost.com/wp-dyn/content/article/2007/09/14/AR2…
I have completed common criteria for a product. It is, for the most part, not very meaningful or rigorous, and hiring the right common criteria team and documentation is key to getting the certification rather than development.
If Opensolaris 2008.05 isn’t meant to ever become certifiable, then I really wonder what in the world Sun is up to with this thing. It certainly doesn’t feel like a “Solaris 11” – that’s for sure.
There are two sides to every story, and as a Contractor who works on a DoD contract, doing anything through GSA is a pain in the a$$. I know this from almost 30 years of first hand experience, not an article in the Washington Post.
Do yourself a favor and stop trolling, because it is evident you do not have a clue.
I’ve worked with contractors like you. Expensive and prone to overstating the whole thing like winning contracts isn’t more about which ex-General and which relationships you have vs. some deep technical requirements.
You simply come across as pompous and all knowing. The easy thing for you to say is simply “look, I know a lot, look at my posts, I work in this field.
Now you are Al-Sahafing the Washington post? “That’s not true, Sun’s relationship with the government is on a bed of roses!”
So now anyone who challenges you in any way is labeled a troll. How convenient. Reminds me of a certain government that calls anyone whose rights they want to violate a terrorist.
Again, if I’m so clueless and you are so all knowing, I have a few issues stated in this thread. Care to solve them? I’m all ears.
If you want to know why people don’t take your site seriously and want to write serious articles for it, you need to look no farther than the comments to this article. The range of comments have went from the serious to the ridiculous. In my opinion mickrussom should have been “kicked to the curb” a long time ago. A person who posts for three hours essentially talking to himself does nothing more than ruin the commentary for everyone else. Nobody that comes to this site should have to routinely separate the wheat from the chaff!
And while I am a few short months from becoming an “empty nester”, I am not inclined to spend free time conducting research and writing what I feel are serious articles only to have something like this happen.
If you want to salvage what is left of your readers, you are going to have to do something in order to dissuade this kind of behavior.
I knew this was coming. Get shown up and now you cry to the editors. I’ve saved everything here, Escue, and it will be reposted elsewhere if its taken down. This not only proves you are a whiner, and cant have a technical debate, but look how you, mister Navy Sir Yes Sir, got an kowtow to the Gestapo Hayden and the TIA and the Government and ask the SUPERIORS to REMOVE THE THREAT.
You cant handle your own, you wan the GOVERNMENT to step in and fix it up for you. I have a 4 inch safety pin and a large diaper for you, what a baby.
Un real. Serisouly. Dont help. Dont prove anything wrong, dont have an intelligent argument. Be pompous, then when you get flak, call in the air strikes.
Were you in the navy when they blew that airliner out of the sky? Or was that before your time? Probably before, because you don’t have a lot of accrued skills from what I can see.
I am Aslan, being put on the tablet for sacrifice, but the table will crack and the deep truth from before the beginning of time will COME OUT and Aslan will be REBORN!
Here is a history lesson for you. Several years ago there was a troll that went by the handle of shaman who announced to the world that OpenSolaris was vaporware. The trollfest that resulted from his comments on various Solaris and Linux related articles made OSNews the laughing stock of the online community. If you look hard enough you can find references to the low opnion of OSNews on sites like Digg, Slashdot and other places.
You are no different than shaman in my opinion, and should be kicked to the curb. Just because people have the right to express themselves doesn’t mean that one should have to listen to your babbling crap at the expense of others who might actually have something to say of value. Your useless diatribe has not contributed to any part of the discussion on OpenSolaris and has more than likely chased off potential readers and commenters, who are sick of trolls taking over and spewing their drivel.
I am a 20 year US Navy veteran, so you can jam your lame attempts at insulting me, for it is you that has consistently proved that you have no business being here!
SXDE 1/08 was way, way better than Solaris 2008.05. Somehow opensolaris went from being vapor, to decent, to crap. I’m not trolling, but I’m not going to put lipstick on a pig.
There you go, revealing yourself to be a well trained Hitler youth. You simply cant ignore me, you must in some way try and force me to be silent. Authoritarian thugs have no place on the internet. You may have us trapped Hayden style in real life, but here we are still somewhat free.
So because you are a vet, you say that proves something. Because being a veteran means that your and your super-class of exalted citizens means there is not equal protection, no 14th amendment, and we need to kneel at the altar of the almighty law enforcement and military goons which are operating illegally per posse comitatus within the borders of the USA. You are part of a standing army which is not the militia and you are illegally enforcing law within the USA.
Next time you swear to defend the constitution, read it, examine how you violate it, and please, work to use it to route out threats both foreign and domestic. I like how the military thinks talk like this is rogue, but the Federal Reserve system and banks can erode the purchasing power of the “dollar” from 2005 to now, stealing from every man woman and child, and you goons sit idly by as the illegal government sells us into being bank slaves, and I, one man with an opinion, needs to be kicked down by these dark authoritarian forces even further, pay taxes, obey all laws, lose my free speech and be enslaved by banks, kick me more. Kick me when I’m already suppressed.
Escue taking orders from herr General Hayden
http://upload.wikimedia.org/wikipedia/commons/1/1b/Michael_Hayden~*…
Show me zee papers!
Edited 2008-06-06 23:22 UTC
I didn’t see any trolling. He did a – harsh, but that’s his right – meta-review and everyone jumped on him for his anti-Sun attitude rather than addressing points he raised.
I was curious so I googled for “shaman osnews” and found this comment in a thread: http://osnews.com/thread?231568
Have to say it rings true from what I’ve seen. Myself and many others commenting I’m sure are not Sun haters. Rather we use Sun products, miss the days when Sun stuff was shiny and high-performance and cutting edge and exciting. Sun seems to be falteringly moving in that direction again which is great, but this hostility toward non-Solaris users or anyone that finds something they personally don’t like in Solaris is getting tiresome.
From the article linked which started that thread – http://www.regdeveloper.co.uk/2007/04/13/readers_feedback_linux_vs_… – here’s a true comment. “Well, one reason that people might choose to miss out on OpenSolaris is because we’re (in general) a conservative lot — once bitten, twice shy — and a lot of people have had bad experiences with Solaris (and, dare we say it, also with Windows and Linux) in the past. No matter how much software and UI improves, it takes ages for the community to accept this. A reputation that took years to build can be lost with one bad release — but won’t be quickly reinstated with one good one. So there will always be people who resist change — and why not, if what they have now works for them.”
Sun would do well to print that comment up as a poster and put it in the hallways. Sun has been hostile to developers since the first days of Solaris, when they threw BSD under the bus and made everything SysV. POSIX_ME_HARDER indeed. That sure worked out well, look how Sun has retained compatibility with the wide variety of Unixes out there while no one uses Linux or FreeBSD because of their oddness.
Sun’s compilers are top-notch and even have a few advantages over GCC in some areas. But they kept it out of the OS for years and years. Now it’s free anyway, what was the point? But Sun people complain that this code or that code only builds in GCC, or somebody linked their package against libgcc. What did they expect? GCC may be a least common denominator but if it’s available everywhere and Sun won’t even let people look at their kit, no one’s going to bother. I’m sure people are trying to retarget Sun’s compiler now but it’s not going to be a high priority and it will take a long time.
Solaris 8 was actually probably the easiest Solaris to build third party stuff against. Because it had nothing. Solaris 9 and 10 include pointless old versions of things to taunt you and screw up the dynamic linker.
X server with no features. No ssh for the longest time and then when it arrived it had no features. Twitchy curses with no features. Inability to even decide what the backspace key does. This may be changing but SunOS has been a desert since > 4.1.4 and perhaps (hopefully) < 5.11. Sun and Sun advocates need to realize this and come to the game with some humility.
Instead we have Sun developers ripping chunks out of third-party software and replacing them with their own code. Their right, but when their own code breaks I have it on good authority that the internal discussion doesn’t focus on “should we have replaced this” and “what did we break” but complaining about the original developer, whose code didn’t break while Sun’s did.
Instead we have Sun mocking Linux for years, Bill Joy taking time out from anti-nanobots rants to mock open source and free software, Sun spreading FUD about the GPL any chance they got, Sun preventing people from even doing their own closed source Java builds for free and thereby providing Sun products to more platforms. Now Sun tries to pretend they have always been #1 open source leader? Please.
Instead we have Sun acting like they are the servant of the Secret Fire, wielder of the flame of POSIX. Any good idea from anywhere else, can’t have it, either because it wasn’t predicted or because Sun didn’t invent it. Well that is a recipe for stagnation and the world moves on, finding new standards. GNU for example: you can build that toolchain anywhere and suddenly the underlying OS is basically irrelevant. I deplore a monoculture and homogenity but much of that is Sun’s fault. And tell me this, if compatibility is so important how come Solaris /bin/sh is so broken, have to look in /usr/xpg to conform; how come SMF won’t read inetd.conf for legacy, even though it reads /etc/rc* just fine?
Instead we have Sun people giving talks about DTrace at conferences and wasting time making fun of Linux for exploring another competing strategy. If Sun’s is so good, just put it out there and let people find it. People found Java despite Sun’s best efforts. Sniping from the side implies insecurity.
Anyway. I have Solaris systems. I’ve liked aspects of Solaris for a long time and always respected the kernel. I hope Solaris 11 when it comes out will meet my needs and those of my users. Sun should focus on retaining and growing its current user base, and on resolving issues from beta users rather than trying to refute them.
There is a difference between shaman and segedunum, I have exchanged comments with shaman and he at least recognizes that some of Sun’s technology is cool. Segedunum on the other hand has nothing nice to say about anything other than Linux. His other unique quality is his ability to pull stuff out of his posterior in order to support his position, an example of which is here:
http://osnews.com/thread?314698
And my response here:
http://osnews.com/thread?314699
His comments on my review aren’t even worth responnding to, and his “problem” with running a python app on an UltraSPARC IV machine and it “magically” running better on a Linux machine is crap. Notice that he has yet to respond with any specific details as to why this was happening. If you have meaningful data that explains the behavior, show us. In their defense, neither shaman or segedunum got personal to my knowledge, like some people here.
Where I work, Solaris x86 got a bad rap because when they last used it in 1997 or 1998 it did not perform well. The poor performance I would attribute to not setting the IDE parameters correctly (if they used Solaris x86 on an IDE system). SCSI performance of Solaris x86 was good, based on my limited use. What bothered me is that nobody made any attempt to periodically check to see if anything improved. The shop I work in has been SPARC for years and will likely remain the same.
As far as compilers go, Sun wasn’t the only company that did not ship a compiler as part of the OS. IBM doesn’t ship one with AIX and one of my friends who used to manage IRIX machines told me that SGI wanted $20,000.00 for their C/C++ compiler. I have no experience with Sun’s compiler, I have always used gcc and for what I need to compile it works just fine. That might change when and if I get Cool Threads servers in the shop, but for now gcc is fine.
Solaris 8 and 9 were OK, I prefer Solaris 10. The only time I had any trouble compiling anything with Solaris 10 is something that was built using either the entire GNU tool chain or specifically made for Linux.
While Sun points at SystemTap as a much lesser DTrace, keep in mind that at one time there was as much rhetoric coming out of RedHat against Sun as there was Sun rhetoric against RedHat. The post where I mentioned the RedHat sales droid is true, this was an actual conversation I had with a RedHat sales rep. And this is not the only example I can quote, it seemed that every meeting I went to where there was a RedHat presence, they always had something bad to say about Sun. I never heard that kind of talk from any Sun employee in face to face meetings.
If I was going to pay attention to a Linux distro, it would be SuSe. All Novell has to do is get out of their IBM OS/2 style marketing (in other words don’t market it at all) and actually push their product. Their management software is clearly superior to anything RedHat has to offer, and while I haven’t priced it out, I wouldn’t be adverse to speaking to Novell formally.
Solaris for me works, but so did AIX and HP-UX. You just have to get familiar with each OS and their quirks, some have more quirks than others.
You know you are too quick to label people anti-Sun.
You seem to think I am anti-Sun, but when Sun canceled x86 Solaris I did find McNealy’s internal email address and did take the time to write to and talk to as many people at Sun as possible to get that thing back.
You simply focus way too much on what “side” people are on, and that is detrimental to Sun to have complete zealots being advocates.
And about the python experiment, hell, if someone just runs “time” on something without profiling the code, and it runs faster, say on Linux, maybe some people don’t have the know how to figure out why its slower.
If every Sun-lovers riposte is to savage someone, say “where is the details” and deny any form of useful help and make a regular Joe feel like a moron for relying on empirical data, then people just go and use something else.
Arun simply did not believe I had found and filed bugs involving ZFS under load. He refused to lend any credence to the issue unless I gave him bugIDs. I know what they are, and could even provide core files. In fact, at one point, the core files I was getting were themselves corrupted, that’s the depth of the issue I was encountering, yet, somehow Arun felt it necessary to simple try and assault those who would find a problem with a Sun product.
First, I don’t have to label anyone or anything as anti-Sun, things just happen. Anyone with more than a few working brain cells that posts here knows that segedunum has nothing nice to say about Sun. FACT. RedHat’s anti-Sun attitiude is almost the stuff of legend, or were you hiding under a rock when the CTO demanded that Sun open source Java, in case you can’t find it, here it is:
http://www.theregister.co.uk/2006/05/18/red_hat_sun_java_license/
I have attended multiple meetings with RedHat sales staff that make it plain as day that they don’t like Sun and waste no time in telling you (minus certain key details) what is wrong with any Sun product.
Also a FACT.
I don’t have to make this stuff up, the truth is so much more fun than fiction. I don’t focus on it at all, you can’t seem to wrap your head around the fact that I call a spade a spade. If you are a troll, you are a troll.
Where you pissed me off:
1. When I asked about using iozone your response was “#2 Don’t tell me how to stress my boxes. Solaris doesnt fail the stress test with UFS, but it does with ZFS, so what the hell does that have to do with IOZONE? I know how to break file systems, thank you, I dont need you to tell me how.”
Do you always bite the hand of the person trying to help you? Strike one.
2. When Arun asked you for the Case ID’s for the cases you submitted to Sun, you could have provided them instead of giving him a lot of static. I couldn’t find your problem on SunSolve (yes I do have a SunSolve account). My guess is you don’t have the “god level access” you claim to have, and you are just whining. Why doesn’t that surprise me. Strike two.
3. One of the clear signs of a troll is when they run out of their comfort zone technologically speaking, they start hurling insults. Anyone who is reading your diatribe is probably wondering whether I have contacted an attorney yet. You clearly are not in the position to judge either my technical competence or the quality of my service to the United States. Strike three.
If I was running OSNews, you would have been kicked out, permanently. Anyone who posts 112 comments and goes on a three hour rant with himself obviuosly needs to get a life at a minimum.
Finally, you want me to fix your silly assed problems, I charge $350.00 an hour, with an 8 hour minimum and require two days per diem up front, a 4 star hotel and first class airfare round trip. I also like a Mercedes or a Porsche as a rental car for the duration of my stay. I also like Cuban cigars, not those cheap Dominican ones, expensive Cognac and large chested blondes (real over silicone). You think your company can afford an “overpaid contractor” who can actually fix problems since it appears you can’t?
Edited 2008-06-07 23:04 UTC
The only way to get kicked out of OSNews is to personally offend an admin. Lesson being? As long as you don’t make fun of Fiona Apple, you can troll with impunity.
That’s funny!
Um, the insults started by Arun WAY before I did and with a much higher frequency. You do have this thread and can read it.
If I’m a roll, then Arun is something way, way worse than that.
And if Arun is advocating for Sun, he is a bloody horrorshow at doing that.
Also, comfort zone. Lol. If you cant gather the level to which I’m comfortable with Solaris by reading the things I’ve said, well, horses and water.
I routinely port things to Solaris and use the Studio toolchain. I refuse where possible to run sunfreeware/blastwave cruft and try to do things the right way.
And again, lets pretend for a moment, that I’m manufacturing the God level access (I’ll be sure to take a picture at Sun in a lab and post it to the web, and be sure while in that Sun building to hold up a current newspaper with your name on it, and pull up the bugster screen on a SunRay to shut you all down, give me a few days.), everything said here is true, there is nothing you or Arun can do to fix it, a simple search on bugs.opensolaris.org for all bugs open and zfs will pull up a plethora of warty bugs with ZFS, and in fact, today I just got a ZFS patch for Solaris 10u5 in the patch manifest.
Whats annoying is getting accused of being a troll when I’m simply reporting facts.
So when you go for the throat and more or less troll me, Ill respond with bombast and the like, since before things were made rude and nasty by Arun, I wasnt getting anything out of the deal, I really dont need you or Arun’s help getting my production systems up, I Have boxes with uptime in the years, so I know how to do that, so what you leave me with is an options to be a bombastic crazy, because being constructive wasnt one the menu for a LONG time now.
Well I’ll say one thing, if nothing else your are consistent and persistent. So, answer this question, what is so difficult about providing the Case ID’s? There’s nothing like shutting someone up by providing them what they ask for. You could have sent it to either Arun or myself through a private message. It doesn’t make any difference whether Arun or I could fix your problems or not, if I wanted somebody off my back and to prove a point I would of gladly given them the Case ID’s.
You mean Arun’s comment to Segedunum, who you don’t seem to think is an anti-Sun troll despite the mountain of evidence to the contrary. In fact there are two posts prior to you even joining the discussion that point out what kind of troll Segedunum is:
http://www.osnews.com/thread?316969
http://www.osnews.com/thread?317006
Or did you miss those on your way to bashing Arun?
You certainly are full of yourself, I wouldn’t exactly call what Steve Christensen and Dennis Clarke provide for the Solaris community as cruft. Maybe you can write an article and show us all the error of our ways and illuminate the path of righteousness along the way while showing us how to compile better software. I won’t hold my breath!!
And I really am looking forward to those photos inside Sun, how about a picture with some of the Sun’s Beta Team, they will know who I am. And on that newspaper list five bugs that I have submitted to Sun on any Solaris Beta. I can check against my archive of messages from each Beta test period.
Toodles!
Like I said, I’ll get to the case IDs. The problem is with you two, is the BugIDs are simply symbolic. They mean so much to you, but nothing to me. They should neither prove nor disprove any given argument. I have resources who work for Sun who, along with my, express frustration. We all like the product quite a bit, but sometimes its hard to effect change. Since I believe my technical credibility is perfectly intact without the bugIDs, your repeated asking for them means that without those, you don’t have much, so the logic that is being employed is that if I don’t have “real” bugIDs, everything I said must not be true. The thing is, to me, the threads stands without any bugIDs yet I can keep a high level of engagement by not providing them. Then at some point in the future, I’ll put them up, and then not only do my logical, effective and true arguments and statements regarding Solaris (and my desire to see it improve) stand, but the bugID mess gets shot down to boot. It’ll great, magnum opus style.
I wanted somebody off my back and to prove a point I would of gladly given them the Case ID’s.
I don’t want you off my back. I honestly feel that the more I say here, the more I will reveal my technical acumen and the more I will corroborate the idea that Sun apologism and jingoism is hurting the overall Solaris effort. You would give CaseIDs simply because you need something like that show how close your are to this inner circle elite. I should say, I’d prefer giving bugIDs rather than a case number – given that a company’s interaction with support is a private matter. I know folks who are into domestic Total Information Awareness (TIA) don’t respect privacy, but unlike those who are well linked into extracting wealth from the taxpayers from the government machine, I need to keep my stream of income because I haven’t exempted myself from the laws of economics.
This is where I bow out, you are so over the top it’s not funny!
I implore you to get off this thread and on with your life. The moment I realized that the best way to deal with a troll is to ignore them I stopped responding to mickrussom.
He is trying to bait you to respond by posting constantly trying to touch a nerve. I almost fell for it once and started to write a response. Nothing feels better than closing a fully written comment to a troll and then simply closing the window and saying “not worth it” and moving on.
I have had a lovely weekend because I did that. I hope you do too.
Thanks! I managed to set up my daughter’s new laptop for college and listen to some thrash metal I haven’t heard before, I’m good to go. As soon as he mentioned TIA, I knew it was time to go!
Gee, I had alluded to it many times before… I guess it takes an overt reference to something for you to follow?
Anyways, I’ll keep tidying up on this thread for a little while. You haven’t been able to debunk or argue anything, so you resort to trying to deconstruct me and throwing mud at me Karl Rove style, and you declare your Porsche Cuban cigar lifestyle like Ghetto Bling, your 20″ Rims on your Escalade, “Man I’m coolin getting paid” bling bling means you are “right?”
Insane. You know the guy who invented the first RISC processor rarely cashed paychecks, he had a passion for work. So if you are coolin gettin paid, it doesn’t mean you are worth anything in terms of technology.
To those interested in the TIA program, please read it here.
It was a proposal (which is still in skunk-work status) to use our international spying apparatus and satellites within the USA – a vast warrant-less dragnet against US citizens.
http://en.wikipedia.org/wiki/Information_Awareness_Office
http://epic.org/privacy/profiling/tia/
Note the “government man’s” immediate reaction to the “TIA,” you are a “moon bat,” “crazy,” etc.
They cannot categorically deny this illegal stuff is going on, so they call anyone who knows about it crazy lunatics, etc.
Feel free to read up on it, you can see that when something causes a “hot poker” reaction like this in a “G-man,” its closer to the truth than they are comfortable with.
SEN. RON WYDEN: Is it correct that when John Poindexter’s program, Operation Total Information Awareness, was closed, that several of Mr. Poindexter’s projects were moved to various intelligence agencies?
JOHN NEGROPONTE: I don’t know the answer to that question.
SEN. RON WYDEN: Do any of the other panel members know this? The press has reported intelligence officials saying that those programs run by Mr. Poindexter – I and others on this panel led the effort to close it. We want to know if Mr. Poindexter’s programs are going on somewhere else. Can anyone answer that? Mr. Mueller?
ROBERT MUELLER: I have no knowledge of that, sir.
SEN. RON WYDEN: Any other panel members?
GEN. MICHAEL HAYDEN: Senator, I would like to answer you in closed session.
Edited 2008-06-08 22:36 UTC
9 June 2008, 12:46
Security hole in Sun Solaris left unpatched for months
Sun Microsystems has reported a vulnerability allowing logged-in users to crash or compromise a Solaris system, some six months after the problem first became apparent. The bug is an off-by-one buffer overflow in the inet_network function in the libsocket and libresolv libraries as well as the libc.so.1.9 and libc.so.2.9 SunOS 4.x binary compatibility libraries in Solaris. The function resolves IP addresses into plain text names and vice versa.
All applications which use the vulnerable library are affected. In principle, the hole may also be exploited remotely if a network application submits parameters entered remotely to the function without further checks.
Sun has so far not provided a fix or suggested a workaround. Other vendors including IBM, Suse, Red Hat and ISC(BIND) have fixed this problem in their own systems some time ago.
Nice job catching those security flaws for Sun mister $350/hr super duper government G-Man beta tester for Sun.
If you are so good, and so worth $350/hr, lets see you give a patch for this flaw:
http://www.heise-online.co.uk/security/Security-updates-for-FreeBSD…
In the words of President Camacho: THAT’S WHAT I THOUGHT.
Edited 2008-06-09 17:02 UTC
And it now stands that nothing I said was disproven or shown to be untrue.
Ad hominems and personal attacks and strawmen, but nothing to refute, just try and discredit Mick instead of doing anything constructive.
Arun, you are cached here, Google, and the thread will be mirrored and shared out. The points about Solaris having issues will probably be taken into account, and the blind religious jingoist apologism and rancid attacks on other people’s ideas and thinking will be here for all time for all to see!
Its win-win-win for me, lose-lose-lose for you, tah tah.
Nothing feels better than closing a fully written comment to a troll and then simply closing the window and saying “not worth it” and moving on.
A fully written comment? I’d like to see that. Seems everything write thus far is fairly light in the well-thought-out department. You should have posted, you might have made just an iota of headway.
9 June 2008, 12:46
Security hole in Sun Solaris left unpatched for months
Sun Microsystems has reported a vulnerability allowing logged-in users to crash or compromise a Solaris system, some six months after the problem first became apparent. The bug is an off-by-one buffer overflow in the inet_network function in the libsocket and libresolv libraries as well as the libc.so.1.9 and libc.so.2.9 SunOS 4.x binary compatibility libraries in Solaris. The function resolves IP addresses into plain text names and vice versa.
All applications which use the vulnerable library are affected. In principle, the hole may also be exploited remotely if a network application submits parameters entered remotely to the function without further checks.
Sun has so far not provided a fix or suggested a workaround. Other vendors including IBM, Suse, Red Hat and ISC(BIND) have fixed this problem in their own systems some time ago.
And it now stands that nothing I said was disproven or shown to be untrue.
Ad hominems and personal attacks and strawmen, but nothing to refute, just try and discredit Mick instead of doing anything constructive.
More to come.
You certainly are full of yourself, I wouldn’t exactly call what Steve Christensen and Dennis Clarke provide for the Solaris community as cruft.
I like to not trust binaries from others as a matter of course, I also like to have a clear documented way from getting to A to B. If I must go outside the Sun provided software, its a serious well documented staged and tested process. I don’t like running to some ftp to grab some random file which may or may not be of the highest quality. I’ve had certain interesting run-ins with the Sunfreeware stuff, but hey, I test and stage things.
The idea that Sunfreeware or Blastwave exists is sad in and of itself. Either FreeBSD ports or Gentoo/Portage coupled with the Studio tool-chains would have been an excellent model in reproduceability, transparency and it would centralize patching for porting to Solaris with Studio so that more and more people can “catch the wave” and get more applications in. Instead we have two fiefdoms, that while a nice service, I believe does not foster porting as well as a model like ports or portage.
BugIDs: 6689906 6642929
What issues are these for?
Thumper.
That’s not an issue.
Well if you can read the bug, you’ll see a wart.
Since all the ZFS bugs are on opensolaris.org, I’d thought I throw you a few that not-just-anyone can make up.
You better hope I can’t pinpoint who you are.
Is that a threat?
Oh, is it that you can’t get to bugster yourself?
I can that’s why I asked.
OMFG, You work for Sun.
Now I can start looking for you.
I have friends at sun.
You do? Because you seem to be doing major damage to Sun’s reputation here.
The bugIDs were flypaper, and you and your cronies took the bait. Now its only a matter of time to find out the extent of this problem.
Is that a threat?
Let’s just say your pals better know someone high up in the foodchain there.
By the way, Sun and RedHat are now in bed together, so the RedHat hate crud should be about over right about now.
http://www.redhat.com/promo/java/
http://www.sun.com/smi/Press/sunflash/2003-05/sunflash.20030519.4.x…
http://news.cnet.com/8301-13580_3-9811395-39.html
Also, I’ll bet redhat gets javaws and the browser plugin working for 64-bit Linux before Sun does, since its been 3 years since 1.5.0 FCS , and they’ve simply ignore thousands of users and developers requesting this be fixed.
Finally, you want me to fix your silly assed problems, I charge $350.00 an hour, with an 8 hour minimum and require two days per diem up front, a 4 star hotel and first class airfare round trip. I also like a Mercedes or a Porsche as a rental car for the duration of my stay. I also like Cuban cigars, not those cheap Dominican ones, expensive Cognac and large chested blondes (real over silicone). You think your company can afford an “overpaid contractor” who can actually fix problems since it appears you can’t?
Tax payers, see where corrupt government contractors get top billing and treatment on your dime. Like McCain, if your daddy was big in the military it brings in cushy contracts, apparently.
Never mind those poor wretched people from New Orleans who still don’t have a place to live, this 20 year veteran needs to charge the taxpayers big bucks to fix things in an OS where the government has contracts to the extent were there any real flaws Sun would contractually obligated to fix them.
Again, the problems I’m having will most likely be resolved, but things shouldn’t have ever gotten this far, you have all of Sun marketing declaring victory WAY before actual victory has been won.
Don’t foist unfinished software as a panacea and you wont get criticized so hard.
Oh, and Cuban cigars are illegal to import to the US. I figure with the kinds of hookups you military types being above the law and all, you are probably exempt.
14th Amendment, whats that equal protection stuff about anyway? You don’t have to know.
If I was running OSNews, you would have been kicked out, permanently. Anyone who posts 112 comments and goes on a three hour rant with himself obviously[sic] needs to get a life at a minimum.
Hitler ran Germany into the ground, too. Except it took him on the order of decades. You running this place? It would be irrelevant in a few months. Most fascist regimes implode.
$350/hr.
This is really rich. Since I am responsible for hires both here and in India, I could have a a team of at least 15 engineers with overhead work on a problem, probably more.
You simply are not worth $350, and if I had a problem that was worth $350/hr to fix, it certainly would be a big job, and I certainly could get the talent to do it far cheaper, with multiple QAs per developer, and far more effectively with a team from India.
You know a lot of Sun software is being made in Bangalore now, don’t you?
Overpaid, arrogant ineffective folks like yourself are really pushing the outsourcing tide, congrats.
Many families in India thank you.
anyone who is reading your diatribe is probably wondering whether I have contacted an attorney yet
Since you claim to be of the government so much, and are mister ex-mil vet, the thing is the first amendment – its a bit from the bill of rights, its protects free speech.
I’ll give you the two clauses that should be sufficient to protect:
Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.
Since I am speaking and not committing libel/slander here, I have the right to free speech. Sorry, what General Hayden probably told you was ok to do, you know, like holding people without bail and sidestepping the Constitution at the CIA – it isn’t the written law yet. But who looks for government guys to follow rules. Your Cuban cigars in point.
The second clause is, that as a super duper government man, I have a right to “petition the Government for a redress of grievances.”
And since you have said I am insane several times, I could always try the insanity defense. I mean, the plaintiff here is feeding the defense.
Edited 2008-06-08 07:57 UTC
FACT: Sun’s bizarre licensing of Java hampered its development, usage, and spread for years.
They refused to let anyone do even closed source ports for ever. So the FreeBSD guys had to beg and plead even to be allowed to make a binary port and release it as “BETA PLEASE HELP US TEST” to FreeBSD beta testers only.
I’ve already mentioned the Blackdown thing so no need to go into it in detail. But I will add that Blackdown had Java ports to Linux on many architectures and when Sun took and re-released their work they released only a “LINUX” binary (meaning, RedHat on x86).
Java could have been on every OS everywhere if they had just loosened their grip and allowed people to help them for free.
From the article:
Are you seriously going to try to claim that wasn’t true at the time that was written?
I don’t trust any sales guy as far as I can throw him.
Are you saying RedHat does these horrible travesties and slander against Sun but Sun is always kind and nice to RedHat/Linux? I doubt it, because I have heard first and second hand of what Sun has had to say about Linux and about GPL and about open source generally until lately.
Sales guys lie, that’s their job sadly. However on the Linux/BSD side I don’t see a lot of advocacy from anyone but distro vendor sales droids. All the developers seem to spend their time writing or fixing code, and the few times they ever proclaim dominance it is always in a self deprecating amusing tone with a smiley 😉 (cf Linus)
Sun people seem to have a lot of free time at their job to talk trash.
I filed three bugs against java to make a jvm for freebsd. They were all rejected. I had to do blackdown and other silly stuff to get jre on there.
It is annoying and its a pain in the butt.
FACT. RedHat’s …
Also a FACT.
head around the fact that
I remember a classic troll with liberal use of the word FACT, see below:
It is official; Netcraft now confirms: *BSD is dying
..<snip>..
Fact: *BSD is dying
I disagree with SuSE, CentOS/RHEL is a far better starting point for enterprise Linux.
I’ve noticed a lot of government agencies using SuSE and when I have to wedge a produce into SuSE, the build environment is really substandard, I have things building on Gentoo, RHEL, CENTOS, Debian/Ubuntu style systems and SuSE will choke. So with SuSE I usually have to make a different make file.
YAST2 is a better all-in-one configuration-station, but it is fraught with issues which makes it easier to deal with if it were broken up into separate utilities that attempted to do less.
I have always used gcc and for what I need to compile it works just fine. That might change when and if I get Cool Threads servers in the shop, but for now gcc is fine.
A guy who charges $350/hr cant even be bothered to use studio. I wonder, does using different compiler than the one that compiler your OS, along with linking to the OS c-library instead of libgcc, well, wouldn’t that potentially affect the OS’s “trusted” or ability to meet certain criteria the government has set forth, I mean, you can just change tool-chians and put random binaries compiled with random un-vetted compilers and still be secure? Sounds haphazard.
That video of the guy trying to hide the flaws in ZFS, pretty damning isnt it? Got under the skin a bit?
Ex military guy folks ex military. Wear you yellow ribbons because boy if they don’t like you they want to take away your freedom of speech.
BTW just wanted to make sure to say I liked your article.
I don’t think anyone found fault with it.
Why argue with people – mickrussom has clearly gone off the deep end – that mention any discontentment with a third-party product? Why not just forward the discussion to Sun so they can review it for their work? Some is probably bad code, some is bad documentation, and some is user error. But all is addressible by Sun and a good opportunity to win customers.
Edited 2008-06-07 01:37 UTC
Thanks.