Slashdot Mirror
Oracle Switching To Linux
Bill Kendrick writes: "This Computerworld story quotes Oracle CEO Larry Ellison as saying 'We'll be on Linux no later than the summer, so we'll be running our whole business on Linux." When asked what this means for Unix vendors like Sun... "It will be several years before the big machine dies, but inevitably the big machine will die.' Ouch!"
I'm sure this is just wishful thinking, but having just gone through the absolutely painful process of getting Oracle to run on RedHat linux, perhaps this move will eliminate the need to use a certain version of a certain distrobution to make it run. *shrugs*
Josh Woodward
They will have to support only some standard distributions of Linux, with no modifications or modifications limited to their "supported" subset they will create.
Otherwise, it will be too big of a hassle figuring out where the problem lies with a custom distribution. This is not really that good of a thing for either Oracle or Linux... because either Oracle will have to have their own distribution, which you can not alter if you want to keep support, or you will have to go with a RedHat, Debian, Mandrake or some other flavor and keep it to their specs...
Interesting to see how this turns out....
Blaming a bunch of "I do it for fun" programmers won't always cut it in big business. Linux and free and open source software are great alternatives, but you always have to weigh the risks against the rewards.
IIRC, the server that Oracle showed off Oracle 8i with was an HP double-wide pedestal server (possibly a PA-RISC server).
E*TRADE Maximizes Efficiencies With Migration to Linux
The problem with Oracle licensing is that on Intel its in $ per MHz whereas on Sparc/PARISC/MIPS/etc. it's $ per CPU. This is what makes the 'big iron' competitive with the smaller machines, paying 25$ per Mhz on a dualie 2.2Ghz P4 is more expensive than paying for a 4 CPU license for an E450.
Couple reasons why Sun still will be preferable to off-the-shelf, commodity Linux boxes for many a year (with or without Oracle's blessing) - and how to change it! (Disclaimer, we run Linux on our web and DB boxen and do NOT use Oracle (Sybase ASE in fact) and were burned by Sun in a deal we wanted for some 280's).
Banks and others with lots of cash have traditionally enjoyed the "Let's buy a couple really really big boxes and replicate them everywhere" mindset and I don't think that will change. Clustering is way cool but I am not convinced the TCO is far less to cause large customers to switch their entire mission-critical, multi-billion dollar a day transactional systems to Linux.
They will stay with what works for a long, long time. Why Larry's pronouncement of 'support' is interesting is that Linux is, for the most part, unsupported. Sun has hundreds (if not more) of engineers around the world on standby -- if your E10K goes down at 4AM they probably know about it before you do (since they have all sorts of neat things built in) and are already on the scene. With Linux? Not so much -- but Oracle is going to try and push the fears of 'what if it goes down at 4am!' out of their minds by saying "That's ok, we can fix it!". Linux and Intel need to offer much of the same features - I know Compaq has neat little remote monitoring cards with their servers, something like that which hooks into Linux and is a commodity (like video cards, or RAID cards, etc.) would help a lot.
Yes, there is an inherent 'single point of failure' with big boxes. That is why they 'cluster' (in name only and not a special type of software) by replicating all their data from their master to several slaves. Currently Sun platform usually has MORE than ample room for growth and you buy 3 E15Ks simply to have warm-standby machines in case the first goes down (and you can always use the other two as readers).
From a TCO standpoint it is far easier, faster, and cheaper to replace a single machine (under warrantee) than it is to have 20 small ones go down at night. Yup - you need to have redundant supplies on hand for the 'worst' situation - and if you have 100 Linux boxes in a nice array and an earthquake hits you now have to order 100 new boxes to replace your destroyed ones. Sun can get you a replacement (or replacements) installed and configured long before the first truckload of new PCs arrives.
Further, you have to configure and maintain 100 boxes vs. a small cluster of Sun machines. I haven't had much experience in large-scale clustered Linux systems but I would surmise that making a kernel change on 100 Linux boxes would take more time and $$ than to 3 Sun machines.
Plus, Sun's 64 bit architecture beats the pants off of Intel -- and in a large DB app you NEED that extra I/O (which is why a 220R with 450MHz x 2 CPUs will spank any dual Intel system out there). I have yet to see any head-to-head comparisons of Itanium and UltraSparc III, so perhaps Intel can rip that from Sun someday.
Thanks,
--
Matt
Ok, i'll bite:
for starters, you are comparing a dual cpu box with a quad cpu box. The quad's always cost more per cpu. Simple matter of the fact that it is harder to get 4 cpus' to talk together compared to only 2 (why do you think intel has yet to produce a 64 way smp server.... sun did it 5 years ago, cray did it before them.) You also have to look at things like backplane contention (are all of those cpu's on the same bus? sucks to be you if they are)
anyway, yes, sun boxes cost more than their intel counterparts in the low to mid range. That said, I have yet to find an intel box that does what the X1 can do for the low end, and once you get to 8 ways systems, intel starts to disappear from the map (and the sun boxes are the same cost or cheaper).
Now, we got the hardware price argument out of the way.
when making a decision, there are 3 major areas to consider: Price, Performance and Reliability. Only an idiot would focus on price when the cost of downtime is a million an hour.
The real reason i purchase sun boxes is not because they are the fastest. You want fast cpu's? Go get an intel box.
here are the main reasons I continue to purchase sun boxes:
#1) Sun's support organization. It is second to none. period, end of story. You have a problem, they fix it. I had a failed disk earlier this week, the support rep's first response was to send a tech on site that day.
#2) When they boast about binary compatibility from $1,000 to $10,000,000, they are not kidding. I can give the developers a low end box and know that the app will still work on a mid to high end box
#3) It just works. I dont get the "what glib are you using", "is that rev XYZ of that nic?" or any of that other crap.
#4) the hardware seems to last forever and ever and ever. And sun supports the stuff for a long time. Every try and get dell to support a six year old box? yeah, good luck.
#5) did i mention the support?
#6) it was built to be managed from a serial port and live on a network from day 1. I love the fact that i can put all of my servers in a colo, walk out, and do the OS install from home. I know that PC's are now beginning to get to the point where you can hook a serial cable up and get them to boot from the net and do an os install. lets face it, there are whole books on how to use jumpstart in the sun environment and do 100% hands off installs. It just works, and it is fully supported.
So, as you can see, there is more to the decision than just cost. In the world that i work in, time is money, and the hardware cost is a very small percentage of the TCO.
I think I know what kernel tweaking you're referring to (which is outlined in an Oracle knowledge base article), and it turns out that if you're running a 2.4 kernel you can just do something like:
/sbin/sysctl -w kernel.shmmax=2147483647 (or whatever you want shmmax to be)
(1) Sun's support is great if you are in the right area. Check with companies in smaller centers to see what kind of support they are getting, and how long it takes to get a good engineer out to resolve any serious issues.
(3) Isn't quite true. The OS is only the foundation, and you rapidly find that you need this particular OS patch for Sybase, another for DB/2, another for Encina, Tuxedo, Websphere, ... If you can find a combination of packages that can agree on patch levels, count yourself lucky! The only advantage Sun has here is a better coordination of patches than standalone Linux.
(4) You have got to be kidding! Sun's CPUs, memory modules, and hard drives fail at least as often as other vendors. Personal experience would indicate IBM and HP as the most reliable, but I have no empirical evidence to support that observation.
Your point on price not being relevant is largely true. The cost of the physical hardware is trivial compared to maintenance staff, software licenses, development costs, and cascading downtime.
I do not fail; I succeed at finding out what does not work.
These comments I could see in this article are the most stupid uninformed balast I've seen in a long time. Maybe its this way for all articles, but I know my ground here and can judge this.
If programs would be read like poetry, most programmers would be Vogons.
Why would you EVER want to use all of those lovely filesystem utilities with a raw device filled with database data? You NEVER touch the database data directly. Want a backup? Have Oracle spit out a backup in your choice of lovely formats, from gigantic SQL statements to reconstruct everything up to your backup software of choice. One uses RAW devices for the same reason one doesn't use SCSI cards with no battery backup; because when Oracle says 'write this to disk' it really really wants it written to disk; not to a disk cache, not to memory, not to a buffer that'll be flushed when the OS decides to flush it. It's all part of the ACID requirements.
Vintage computer games and RPG books available. Email me if you're interested.
Let me guess, you have a multimilion contract and they had a newbey which could use some training :-)...
#2) When they boast about binary compatibility frSo, as you can see, there is more to the decision than just cost. In the world that i work in, time is money, and the hardware cost is a very small percentage of the TCO.om $1,000 to $10,000,000, they are not kidding. I can give the developers a low end box and know that the app will still work on a mid to high end box
eeuh, 2 posibilities here : a) you're talking hw, and i don't understand you at all. The only one which did care for downward campatibilit was INTEL (also only reason why it stayed popular) b) you're talking about software and then it's just stupid. Just recompiling you're app for newer hardware gives you a better performing app. Binary compatibility is just a stinky way to be able to hide theire source.
#3) It just works. I dont get the "what glib are you using", "is that rev XYZ of that nic?" or any of that other crap.
#4) the hardware seems to last forever and ever and ever. And sun supports the stuff for a long time. Every try and get dell to support a six year old box? yeah, good luck.
Right, but for SUN's 1x price I can get a a newer box each year.
#5) did i mention the support?
Euch you mean the part where you get forwarded from helpdesk to helpdesk and finaly get a ticketnumber saying you're in their problem database ???
#6) it was built to be managed from a serial port and live on a network from day 1. I love the fact that i can put all of my servers in a colo, walk out, and do the OS install from home. I know that PC's are now beginning to get to the point where you can hook a serial cable up and get them to boot from the net and do an os install. lets face it, there are whole books on how to use jumpstart in the sun environment and do 100% hands off installs. It just works, and it is fully supported.
Correct yourself here too.. you are talking about the UNIX way, not about the SUN way. The same can easily and much cheaper be achieved on PC hw with a free unix like bsd or linux.
So, as you can see, there is more to the decision than just cost. In the world that i work in, time is money, and the hardware cost is a very small percentage of the TCO.
Please stop glorifying SUN, the only reason you need them is because they have an IT department with a legal department to back them up. (which is the key for most of their businesses) For the rest it's just a big corp not much diffrent from M$ : some brilliant guys and lot's of morron's acting important
--red
if an instance goes down, you dont loose that data
I suspect that in the event of an instance going down, you would not "let loose or release" the data. However, it is possible that you would fail to retain it. The word you were looking for is lose.
Congratulations! You have been participant #18 in my campaign to rid Slashdot of this error.
Proudly correcting Slashdot's most irritating linguistic error since 2002.