Slashdot Mirror
Linux to Replace Solaris at Duke
wwhsgrad2002 writes "At the end of the 2004-2005 academic year, the Sun Solaris computers available in public computing labs at Duke University will be replaced. The replacement computers in these spaces will be Dells, running a version of Centos 3.3 as supported by Linux@DUKE. Pragmatic and technical considerations have driven this change, as Linux continues to gain a greater userbase and more third-party commercial software is made available on the platform. Are other universities eliminating Solaris in favor of a Linux distribution?"
But my company is moving away from Solaris because the new Dell Boxes are at least three times as fast as the fastest Sun we have.
And cost one third as much!
Raydude
Not that the CentOS distro is bad, but it's really more for a server, not a user box. Since this is going in the computing labs, and presumably the students will be logging into the box(es), it would seem to me that using another distro more geared towards users would be appropriate, since the CentOS 3.3 is geared towards enterprise servers.
I'm sure it can be tweaked to be just fine, but it seems kind of an odd choice to me, for a computing lab.
The math department at University of Maryland, College Park recently decided to replace it's Sun workstations with linux computers, probably Dell's.
I for one welcome our Educational Linux using ahchchhc cough cough.
BYU switched several years ago. By the time I took CS 240 back in 2000 what had once been the UNIX lab was full of Dell linux boxes.
http://nerdfortress.com/
Really, so that means vendors have stopped supplying new softwares for Solaris! Or does it mean that practically Solaris is not technically a viable solution?
I really don't see the need to replace an X system with Y system when the X system does the job for you more than adequately. I don't understand why people are always eager to change systems. Of course someone is going to reply to me and say - "hey universities are research institutions and they need new stuff" - too overrated. I am not trying to root for Solaris here, just don't get why you need to replace a system that can do the job that Linux can.
Some companies have said that if Sun was doing three years ago what they are doing now (Solaris 10, OpenSolaris, free licensing), they would not have switched to Linux. Consider that Sun still guarantees binary and source compatibility when migrating to Solaris 10 from older versions, while Linux cannot. Linux is very useful, but there are still things that make long-term deployments awkward at times. Mod what you will, but it is true.
-- Microsoft is the most expensive commodity operating system and office suite vendor in the marketplace.
Hmm, 1/3 the cost, 1/2 the longevity.
Sounds like a good deal to me!
Where I teach, the tech people are linux-phobic. They are adamant about "keeping linux off the network" yet aren't so pissy about OS X (which probably means they've been reading Gartner). Of course, the highlight was a few years ago when I was running linux my older laptop, surfing the net, and doing my grades (through wine no less), and the school's distrtict tech guy asks how I can do this since "novell doesn't support linux." I guess our network admin never heard of, what's that thingy called? oh yeah, TCP/IP.
My problem? I was perfectly gruntled, until some numbnuts came by and dissed me.
Sounds like a good deal to me!
I think you're trying to be funny, but it is a good deal - if you buy two in a year instead of one, each at 1/3 price, you pay 2/3 the price - thus saving 1/3 the price. Since failure is unpredictable even in expensive equipment, you're going to buy two of your servers for redundancy anyways (right?) - so the longevity argument doesn't even factor in.
In three days time I will no longer work for Sun since I have been made redundant.
During my time at Sun I was part of the Companion CD team. We built on x86 and SPARC. For x86 builds we had a Dell 6400, Dell 6600 and finally a Sun V40z (4-way Opteron 246). For SPARC we built on E450, E4500, and V880 (8x900MHz UltraSPARC III) and V880 (8x1200MHz UltraSPARC III).
Now, I will not go into a long spiel about the realtive merits of the various hardware platforms, and I have no axe to grind now since I get my lasy pay cheque in a fortnight but:
Why the heck are you buying (32-bit intel) Dells when you can buy (cheaper and faster 64-bit) Opteron boxes from Sun? If you are a Linux fanboy, Sun will sell you one with DeadRat or SuSE. They are Windoze certified in case you have had a lobotomy, and you can run the free (as in beer) 64-bit Solaris 10 on them.
pBut hey, it's cool to hate Sun on slashdot.
Stick Men
I'm one of the two people here at UMBC who run the core servers for the campus.
;)
We use AFS here for everyone's home directory, mail spool, web space, and other things. To maintain this, we currently have about 6 servers with direct-attached storage serving everyone's AFS home directory volumes. These servers are a mix of Dell and Sun gear running Linux and Solaris. Both platforms have run well over the years, but each server's direct-attached SCSI storage is limitting and, well, aging.
So we can better use our storage and improve things for everyone in general, I'm in the process of rolling out a fiber channel SAN with new servers and RAID arrays to replace what's currently running. The new server gear we chose? Sun's V20z Opteron server running Solaris 10 . Linux is right out.
Why no more Linux, or rather, why Solaris? A few reasons. Solaris's storage management is TONS easier to deal with and do interesting things with than what is available in Linux. Namely, we've found and have been fustrated by Linux's software RAID. Yeah, it works... but that's about it. Weee look, I can make a mirror! Solaris's SVM (aka DiskSuite) is no VxVM, but it does allow us to do things such as disk sets to share between hosts and monitor our metadevices in detail. Linux's raidutils on the other hand are poorly documented and toublesome (usage options don't match reality, etc)
Another aspect on Linux vs. Solaris in mass storage is (as far as I know) a lack of multi-pathing in Linux. Multi-pathing is a no-brainer especially in the context of Fiber Channel networks and Solaris's MPxIO is in-built and works quite well.
But I'm just poo-pooing Linux here on this specific point. We offer Linux workstations in every one of our computing labs. Linux replaced SGI/IRIX workstations there many moons ago and work well for that purpose. Linux servers also are used for our general shell login servers. But on the backend, where we need reliable features, consistency, and heavy-lifting... we're enthralled with Sun x86 servers and Solaris 10. The V20z Opteron hardware actually is cheaper (for us) than a Dell 2650 and offers a ton more features all-alround.
There is an irony, though. The service processor on the Sun V20zs run Linux. Ah well
do students massively prefer the PC's to the sunblades and sunrays? sure. many professors care less. but do we want to limit any of them to a single platform? definitely not.
Dell=cheap crappy hardware. We have 1600 dell workstations (Optiplex GX1-GX280.) Each month we replace 30 - 35 Motherboards that have failed. We never had this problem with the older Dells(GX1-GX110). We also got a bad batch of maxtor hard drives that have had about a 70 percent failure rate in our Dells. Most problems have been with the GX270 line. Out of our first 25 GX280's we have already had 1 MotherBoard failure and 1 Hard Drive. Dell has admited that they have had some problems and sent us 10 motherboard to keep on hand.(Some days we replace 5-6 motherboards) Most of our PC's are used 24/7. I am actually a Network guy buy since our Netware servers never go down I help out with the Dell hardware replacements.(we do NOT use Dells for servers) We were going to switch to IBM Desktops which in my opinion are much better than the Dell's but after IBM sold their desktops to Lenovo we sent all of our IBM's back.
So I'm going to burn through twice as much power to move zeros around that will never be used?
Somehow I doubt that a doubling of pointer widths is going to result in a doubling of your power requirements.
General purpose computing doesn't need to deal with over 4 billion unique things.
Yes it does, all the time. Not all of us write webapps all day. I work in bioinformatics and hit my head on the 4 GB memory limit constantly. There are 300 billion bases in the human genome, and tens of millions of polymorphisms with information required about their names, aliases, positions, and allele frequencies. I can't store things as first class objects- I have to use RLE encoded primitives everywhere and there is no type safety because everything has to be an int. Many algorithms require repeated visits to arbitrary points on a chromosome so paging through a database is not really an option. If you have to page contigs in and out of memory, many genetic linkage algorithms will take the lifetime of the universe to complete.
The 32->64 bit problem isn't the same as the 8->16 or 16->32 problem. If it was, why not just jump to 128 bit?
The Earth weighs 6E24 kg. 0.375% of it is continental crust, roughly 15% of which is silicon. If you consider that the atomic weight of silicon is 28 g/mol and figure roughly 10000 atoms of silicon per bit, that means that if we were to mine all of the silicon out of the continents, make RAM out of all of it, and put all that RAM in one big giant computer, that computer would need to be designed with an address space 132 bits wide. So you see, even 128 bits is not enough.