Red Hat’s announcement that it is adding network and system monitoring to Red Hat Network isn’t that much of a surprise. More unexpected is Red Hat’s decision to support Solaris with it. In other news, the Fedora Foundation has been delayed.
Red Hat’s announcement that it is adding network and system monitoring to Red Hat Network isn’t that much of a surprise. More unexpected is Red Hat’s decision to support Solaris with it. In other news, the Fedora Foundation has been delayed.
If you are going to get into the system/network monitoring market and successfully compete with IBM (Tivoli), Computer Associates (Unicenter TNG), Hewlett Packard (OpenView) then you had better have the ability to support a homogenous environment which includes Solaris. Since RedHat has significantly stepped up its sales effort to Government customers (a big Sun customer), having no support for Solaris could easily be a “show stopper”.
I adore Fedora. It is my favorite distro. I have huge respect for Red Hat. They have gone well beyond any legal requirements to be a good citizen of the open-source comunity. BUT… this foot dragging is getting tiresome. When it comes to matters involving RHEL, the company can turn on a dime, move with lightning speed, and would probably qualify for an Olympic medal if they awarded them to companies.
But in matters concerning Fedora… well, if a company could legally qualify as being physically handicapped, Redhat would be a shoe-in. Delays… excuses… appologies… followed by more delays. There was more employee time devoted to appologizing for taking years to set up the CVS server (and not having it done yet) than most people would spend actually setting up a CVS server. You can bet that if a foundation needed to be set up to further RHEL’s goals, it would have been done with expedition.
Look, RedHat. If you are going to set Fedora free, just freaking do it. And if you are not, just fess up and say you are not, and that you prefer to keep control over it, which is what you really want to do anyway so that it will serve RHEL’s needs better. Make up your mind and act decisively, please.
Not doing so just makes you look a bit sleazy. And Red Hat, Inc. has higher standards than that, right?
You are a retard. Faildora whore lunix sucks. GNUbian Deb/Lumarckes is the way to go… or OpenSolaris, if Solaris 10 puts you 10 moves ahead of the competition, OpenSolaris puts 10 moves on your mom every day.
Yes, I get your point.. but what about Redhat prices??
Have you tried Sun Update Manager Notification/Sun Update Management lately?
No we havent’t (military site), so the tools provided by either RedHat or Sun would probably not be used to update machines remotely.
As for its Solaris support, Red Hat sees this cross-platform support as being helpful for Solaris users migrating to RHEL.
I just don’t see that many shops migrating from Solaris to RedHat. Who would be foolish enough to do that?
Our shop (major EDA firm) used to use Solaris/Sparc as our primary development platform. Now we use RHEL (we also support SuSE). Our customers (pretty nearly every chip maker in the world) mainly use Linux these days. Solaris/x86 was revived too late to make a difference, it isn’t much used in EDA because so few tools have been ported to it. It’s much the same for our major competitors.
No one does large chips with Windows, by the way. It’s always been Unix/Linux.
Sun made a bad mistake years back when they tried to kill off Solaris/x86 to protect their Sparc business.
Sun attemtpted to kill off Solaris x86 because some thought there was very little demand for it. These people wanted to take the personnel and the money spent and put it elsewhere. Obviously you have never heard of the “Secret Six”, the group of Solaris x86 users that met with senior Sun management and helped to bring back Solaris x86. It had nothing to do with protecting the SPARC line at all.
And considering all of the effort that Sun is putting into Opteron machines (V20z, V40z, etc.) I don’t think Solaris x86 is going to die anytime soon.
And considering all of the effort that Sun is putting into Opteron machines (V20z, V40z, etc.) I don’t think Solaris x86 is going to die anytime soon.
Not to mention Solaris X86 has always been a better choice then Linux has for server side applications.
the old solaris x86 sucked… the new one is badass….
so saying that the older earlier vesions of solaris x86 sucked is true.
i dont think linux can compete with the new version
“I just don’t see that many shops migrating from Solaris to RedHat. Who would be foolish enough to do that?”
Obviously, lot’s of people, and they’re actually not foolish at all. Just look at how Linux all but destroyed Sun’s cash cow Solaris/Sparc/big iron server business, and a very substantial part of that was Red Hat. Linux ate Solaris’ lunch by offering a much, much better value proposition by being both cheaper and by running on much cheaper hardware (x86 as opposed to Sparc), while offering comparable performance/stability/scalability.
Yes, Now Sun has a whole new value proposition. Solaris 10 now runs well on x86 (older versions did not). Solaris is free, with for pay support/services (just like Red Hat). Solaris is also arguably the most capable “big iron” OS out there. And now Solaris is Open Source, but with the CDDL, and under Sun’s control.
But Sun was too slow to get to this point. The market already move ahead. And even though Solaris is now free as in beer, as well as open source, RHEL (along with SuSE) remain more attractive due to less lock-in, more hardware support, larger community, more software available, etc.
This is not a knock on Sun or Solaris. It’s just pointing out market realities.
But Sun was too slow to get to this point. The market already move ahead. And even though Solaris is now free as in beer, as well as open source, RHEL (along with SuSE) remain more attractive due to less lock-in, more hardware support, larger community, more software available, etc.
Solaris has the paid support infrastructure and a proven track record that any version of Linux lacks the two key areas when considering enterprise solutions. Hardware support is only a factor on the desktop side of things. I would disagree that there is more software for RedHat/Suse vs Solaris smal rinky dink applications for sure but critical big business applications are more prevalent on Solaris then Linux. Personally speaking I still would never recommend Linux over Unix at this point.
Solaris x86 prior to Solaris 8 blew nasty goats; it took a combination of luck, skill, and voodoo to get those versions (especially 2.6) to install on anything that differed in the SLIGHTEST from the official hardware compatibility list – and even that wasn’t an ironclad guarantee.
Solaris 8 was a big step up but, compared to Linux at the time, it was still SLOWlaris and was behind in stability.
Solaris 9 was better for speed and stability but still trailed Linux by a pretty wide margin; the hardware compatibility was still not even in the same plane of existence as Linux but, if you stuck to the HCL, you could make a case for either one or the other on x86 depending on the apps you needed to run and the inclinations of your admins.
While Solaris 10 has some cool technologies, it’s still
quite new so I’d be leary of putting anything mission-critical on it.
Maybe when it stabilized more, Linux might have something to worry seriously worry about but judging by the average time it takes for well-supported, actively developed operating systems to stabilize, that’s a couple of years off.
The term “slowaris” was first used in 1993 comparing Solaris 1 to SunOS 4.x, I found this out in a discussion on the OpenSolaris forums. It seems to have been used ever since to blast Solaris x86 (and in some cases SPARC) performance as compared to a variety of operating systems (usually VMS and Linux).
The hardware support horse has been beaten to death, let’s move on.
I cannot speak for Solaris 2.6 on x86, but I can speak for Solaris 7 onwards. If you follow the guidance on the Solaris x86 FAQ (http://www.sun.drydog.com/faq/6.html) concerning maxpgio (Solaris 7), and the two IDE parameter changes (Solaris 7 through 9) the disk performance improvement is significant. This is basically no different than using hdparm to modify the characteristics on Linux (http://www.linuxdevcenter.com/pub/a/linux/2000/06/29/hdparm.html?pa….) Just as Linux has improved in its use of Ultra DMA devices, so has Solaris.
As far as deploying Solaris 10, there are people still using Beta builds “in production”, I think that says a lot about the quality of Solaris 10. There is no such thing as a risk free deployment regarless of what OS you are using. At some point you have to decide to take the risk, some people took it to run Linux over another OS. Either take the risk, or don’t.
And finally for JeffS, I live in an area that has a large base of both Sun and IBM “big iron” and I don’t see them dropping their pSeries (AIX) and SunFires (Solaris) to go to Linux. And I also don’t see big iron going for next to nothing on eBay either. If what you said is true, then IBM and HP would not be feeling it as well as Sun, and they are. That is why HP is marketing the Integrity line of servers to satisfy the needs of those who run Windows, Linux, or HP-UX, and IBM puts a great deal of effort into getting Linux to run on their hardware. So I think everybody has taken a hit with Linux, I just think the media harps on Sun a lot more than the others.
There is an article on DBAZine comparing a Linux cluster to a Sun E4500 (http://www.dbazine.com/olc/olc-articles/ault8/view?searchterm=linux)
And there is this entry in John Brady’s blog called “RAC – Many small or a few large nodes?”, pay attention to the paragraph about how Oracle uses RAC.
I’m well aware of the Drydog FAQ; it’s helped me solve a lot of problems with Solaris and as far as perspective goes, not only did i pay for the early Solaris x86 versions, I was listed on a petition to save Solaris x86 when Sun first announced that they would deep six it. I also registered for a license for their community source release of Solaris 8.
That being said: there’s a good reason why DMA had to be enabled for Solaris 7 – it’s was flaky. I tried 5 different hard drives by 3 different manufacturers and they were ALL unstable with DMA enabled. It may have been the fault of the IDE controller but I switched motherboards as well to no effect.
In contrast, hdparm modifications has never locked up anything of the 8-odd DESKTOP machines I’ve installed Linux on although one LAPTOP didn’t fare so well.
And it took Sun a LONG time to figure out how to efficiently move data off of an ATAPI CDROM.
And, to flog the hardware support horse one more time, why can Linux, whose developers have had to scrape, hack and steal device info to write drivers, support so many devices so well when SUN has done so poorly in that arena despite their resources and industry clout.
Finally, Solaris has the dubious distinction of being the ONLY OS that has ever dumped core because of a badly written vendor-supplied network driver – the machine would perpetually reboot. No other x86 OS I’ve ever used, no Linux distro, not xBSD, Windows, DOS, BeOS, nor QNX has ever failed in such spectacular fashion.
To move this discussion back on topic… I think it’s a great move that the company that Sun hates more than any other in the world, Red Hat, is responding with the opposite message. By supporting Solaris, they are telling prospective customers that although Sun won’t jump through hoops to play nice with Linux/RH (is Project Janus dead or alive this week?), Red Hat wants to support the broadest range of environments possible, including Sun. It sends the message that Sun has something to lose in competiting with Red Hat, whereas Red Hat doesn’t seem phased by competition with Sun.
Read the first comment, RedHat has no choice but to support Solaris. If you are trying to market an enterprise grade monitoring solution that doesn’t handle Solaris, you would get laughed out of the office. And since the vast majority of software that runs on Linux also runs on Solaris, what difference does it make whether Project Janus goes anywhere or not. I would much rather see ZFS than Janus.
I don’t think Sun has that much to worry about in regards to RedHat. For all those people who believe that a bunch of x86 boxes are better than big iron, let them waste their time and money. Sun and IBM will be around long after that idea has died.