Dirty little sekret
Neteeza is getting blown away by this thing called Hadoop....
IBM is spoiling for a fight with Oracle, as was abundantly in evidence at Big Blue's recent Investor Day at the company's TJ Watson Research Center north of New York City. Thank heavens – otherwise server junkies might die of boredom. IBM doesn't talk as much smack as Oracle does – or Sun Microsystems did - but that doesn't …
yeah, sure, you can use the "high business value" off the shelf. One problem - your competitor uses the same so you all end up giving business value to the vendor and gaining NIL competitive advantage.
Alternatively you may chose to actually to risk trying to _GAIN_ a competitive advantage in a specific area. Then you find developers and get ha-dooped. Or whatever else you chose to use.
It is a matter of strategy. When all is said and done a company is out there to compete and grow. One that always choses the "high business value" pretty much openly declares that its intention is not to. Pity none of them actually tells that to shareholders. IMO they should (or someone else should).
"There is a huge difference between open source technology and enterprise appliance offerings that provide high business value"
I suggest you do your homework on the topic... You're right, there is a huge difference. In terms of TCO, The Hadoop ecosystem wins out. IBM's core customers are already seeing this and are looking at PoC's using Hadoop.
IBM already sees what Hadoop is doing to their business, and after their disastrous initial foray in to Hadoop, they've made changes to their approach.
Neteeza is losing their lunch to Hadoop.
IBM seems to be so laser-focused on Oracle's slow SPARC and commodity systems that they're ignoring what is probably the highest-scaling general-purpose server - the HP NonStop. Just because HP is too stupid to market it anywhere outside its happy little high-margin telecom niche doesn't mean it doesn't have potential.
.... because Oracle are playing dirty and targeting IBM customers with Exadata solutions, though hiding the costs or that fact that you can't buy the hardware they used to produce their performance data. HP are being the usual HP and not only trying to break in to new markets but upsetting existing clients by refusing to give discounts. Now if HP join in the trumpet blowing then maybe IBM will target them with their blurb as well.
Is it really true? According to IBM, AIX is the most reliable Unix out there. According to IBM, Mainframes are the most reliable big servers out there - and also fastest in the world ("z196 cpu is the world's fastest cpu" says IBM in big press releases).
Awkward it is, that we read about the TurboHercules Mainframe emulator that gives 3.200 MIPS on an 8-way x86 server - that is quite decent performance for a fraction of the price. TurboHercules says their customers are interested, because when their Mainframe crashes, they can switch to an cheap x86 server. Why is IBM sueing TurboHercules for this - when "Mainframe never crashes"? And, recently, in a large IT-newspaper we could read about "their IBM Mainframe never crashes - last time it crashed was 6 years ago".
And the "worlds fastest cpu", z196, is in fact slower than an Intel Nehalem-EX. You need several z196 to match one Intel Nehalem-EX in terms of cpu power. The Mainframe cpus are slow, but their strength is their reliability. RAS costs a lot. Performance is cheap. That is the reason the Mainframes never crashes according to IBM - they have high RAS.
And about the "slow" SPARC cpus, sure they are slow, says IBM. Let me see, who has the world record in TPC-C? 30 million? Who has several other performance world records, beating the POWER7? It is funny that a 1.6GHz cpu can be fastest in the world on several benchmarks, even faster than the "mighty" POWER7. Just read them benches and you will see it is true.
But hey, IBM claimed that IBM still had the TPC-C world record, because "the IBM cores are faster than a Niagara core" - so I would not be surprised if IBM keeps telling everybody they have the TPC-C world record. IBM better pray their customers do not check those "IBM facts", for instance, goes to the TPC-C site. Some would call it pure IBM fud.
That's quite a mouthful, Kebabbert. I'm going to try to address the silliness here bit by bit.
"Awkward it is, that we read about the TurboHercules Mainframe emulator that gives 3.200 MIPS on an 8-way x86 server - that is quite decent performance for a fraction of the price."
Wow! Roughly 6% of the CPU performance of a mainframe, a system designed for moderate CPU performance and high I/O! Truly impressive!
"TurboHercules says their customers are interested, because when their Mainframe crashes, they can switch to an cheap x86 server."
I've spent a fair amount of time with TurboHerc marketing materials, and they don't claim this. They say that a commodity system can be a warm standby failover system. Nobody, even NonStop or Stratus customers, puts all of their eggs in one basket when they can avoid it.
"And the "worlds fastest cpu", z196, is in fact slower than an Intel Nehalem-EX. You need several z196 to match one Intel Nehalem-EX in terms of cpu power."
Floating-point, maybe. Integer, not close. I/O bandwidth, orders of magnitude of difference.
"And about the "slow" SPARC cpus, sure they are slow, says IBM. Let me see, who has the world record in TPC-C? 30 million? Who has several other performance world records, beating the POWER7? It is funny that a 1.6GHz cpu can be fastest in the world on several benchmarks, even faster than the "mighty" POWER7."
TPC-C is a meaningless benchmark, and has been for years. Sun wisely abstained from it, recognizing that it was basically silly. Additionally, I've been hearing some really interesting things about the supposedly invincible T3, although I haven't used one myself...
-Oracle reps taking months to provide a price quote, resulting in customers having to compare Oracle list price to HP/IBM discounted price
-Performance on throughput-oriented multithread workloads being vastly slower than a cheaper AMD Magny-Cours
-Failure by Oracle to provide credible roadmaps for >4socket machines
Anyway, if you insist on using TPC, the fastest non-clustered result is a Power6 machine, followed by a Superdome. If you do some math, it can be assumed that a single Power 795 would be well into the 15mnTPC territory. How many boxes did it take Oracle to get to 30mn again? And how many licenses are they selling, for that number of cores?
We have a very very slow memory leak where network memory buffers do not get released, takes around 9 months to hit the "wall" which is the limit before AIX stops allowing mbufs to be allocated. a 6.1 TL4 upgrade fix's it but upgrading isn't easy to do until the app stack is tested with it. We are running Veritas Global Cluster and my guess is it is either the local heartbeats or the remote heartbeat to the remote cluster. Are you running HACMP?
AIX has always been more prone to memory leaks than Solaris in my opinion.
Biting the hand that feeds IT © 1998–2019