Guru: Oracle's storage strategy is 'screwy' Oracle's storage range is over-priced and incomplete, and its strategy is skewed by being too database appliance-centric, according to a prominent storage guru. Enrico Signoretti, the CEO of Italian storage consultancy Cinetica, has earned a wide following in the storage world through his blogging and tweets (@esignoretti). …
Enrico Signoretti, the CEO of Italian storage consultancy Cinetica, "I don't think they have a real storage strategy at the moment. They need a high-end storage array but there are no companies left to buy. They need to improve the ZFS-based 7000... They need a better price for the 6000/2500"
I think everyone agrees Oracle needs some better pricing on the low end... or perhaps a low-end disk storage array.
I am uncertain what exactly the gaps are and what needs improvement. The platform capabilities are substantial:
- dedup
- compression
- encryption
- replication
- practically unlimited lun size
- practically unlimited snapshots
- practically unlimited quantity of drives
- sata & sas drive support
- fiberchannel server support
- iscsi server support
- infiniband server support
- nfs fileshare support (with transparent identical cifs permissions)
- cifs fileshare support (with transparent identical nfs permissions)
- pass-through support for replacement of old storage arrays for zero downtime transition
- build-in performance analytics
I don't think I have seen better performance analytics in a competing disk array on the market... at any price.
Honestly, I don't know what other feature is missing... does another storage vendor has all the features in a single product???
The issue here is not 'features' on paper, but actual working features in practice!
As Enrico pointed out, the 7000 series has been very bug ridden since its (premature) release and Oracle seem to have cut and/or aggravated a lot of the key developers behind ZFS/7000 series so they appear to be headless now.
We know from experience how poor the 7000 has been at doing what it should have done, and how inept Sun were at fixing it, and how much worse the system has become now that Oracle has merged the support procedures.
It is sad, not just as an end customer but as an observer of technology, to see what should have been a game-changing system that could have wiped the floor with a number of incumbent suppliers being lost through basically bad project and support management.
p.s. The claim of "nfs fileshare support (with transparent identical cifs permissions)" is not supported with NIS authentication for sure, maybe only with ActiveDirectory, so it is less than meets the eye to any Linux/Unix-leaning organisation like us.
David,
I sold it in the past and customers are still very angry.
There is a big difference between a feature and an usable feature! 7000 is based on the latest opensolaris builds and many features remain buggy for a log time after the release.
In my personal experience I saw a very long history of customer waiting for critical patches on CIFS (more than a year!) to use it in production.
about analytics you can find good/better features in Compellent or other next generation arrays.
And what about the future development?
Have you seen the Oracle update from 7X10 to 7X20? it's a shame: i.e. Customers with a 7410 can't migrate and no more expansion trays for them, it's ridiculous.
In theory 7000 is a good product but in practice it is still very immature, with bad support and pricey... and we don't leave in a theoretical world.
This post has been deleted by its author
Enrico Signoretti seems to have forgotten about Pillar Data. In fact, I'm surprised Pillar hasn't already been borged into the Oracle hive, but then maybe Larry planned to sort out the Sun mess before adding Pillar. If he does, Larry gets to give himself a massive pay-off at the Oracle stockholders' expense.
well, who is Pillar? :-)
Pillar is no more credible and a very noisy silence coming for that company (IMHO).
Oracle is a public company, I don't think Pillar is a good acquisition for them. Are you sure that Oracle's stockholders will be happy about such an acquisition like this one?
From the products point of view Pillar hasn't the right product, They can't compare with 3Par,HDS VPS or VMax.
Whilst it would be true to say the current generation of Pillar Axiom products are aimed at the midrange (which means they already compete with the lower end of the HDS, EMC and 3Par enterprise offerings), their storage pooling technology and automated tiered manangement could be of interest to those customers looking at replacing monolithic top-end arrays. At the worst, Pillar would make a good fallback plan for Larry, and they already have a customer base to build on so not such a bad investment for the Oracle shareholders.
We are a Pillar customer for three years now...yes, they are not as big as the other ones and we didn't use them in the high-end environment. But the performance and utilization we got out of these boxes using SATA drives was incredible and even better than our high-end environments. And now our vMax is having a hard time to get the same performance I can tell you. Pillar's problem in the past was the code stability, the disruptive code upgrades when installing a major release and enterprise support (at least here in Switzerland) . those are the points where Pillar cannot/could not cope with high end companies/products..
But I don't want to forget to mention the ease of management compared to the vMax.....and the enhanced support for Oracle's ASM (stripesize...).
Even if Oracle's HW support at the moment ist the worst you can get (just pure crap), they are a big player and present everywhere which would help Pillar in the support issue.
So if Oracle could position it right, Pillar could be a better acquisition than you think. And I know people at Pillar...do you?
Oracle's storage offerings are even more limited than the article suggests - as far as I'm aware the 2500 series is no longer offered.
Equally damaging is the removal of the X4500/X4540 line, which didn't really have much competition. There's no direct attach storage either. This is particularly odd given how well such storage plays with ZFS.
They were fine before buying Sun .. but the question i'm asking , because they became a very gaffe prone company that seem to lack clear direction is this : did they become too big too fast ?
As of late they seem to pi** everyone off and generally drive every developer bonkers.
Java , Hudson , Open Office , Open Solaris .. Where will it stop ?
Seems like everything that had community value is going to be forked and everyone turn their
backs at that " has been " company. Yes Larry , you're one of them " has been " too :)
That felt sooooooooooooo good ..
The fact that Oracle insist on still talking about tape says all I need to know. The rest of the world went disk based backup yonks ago and those that haven't yet are seriously contemplating it. Use tape for offsite permanent archiving and audit/compliance fair enough, but rapid availability DataDomain et. al are running the show and Oracle had better get down the jetty pretty fast if they don't want to miss the boat!
You need to check out the history of Oracle. Larry and company have performed consistently well over the years. They are able to maintain nearly a 40% margin on revenues. Oracle does this in spite of it's customer base. Larry focuses on the 20% of the clients that bring him 80% of his revenue.
And he hires really smart people to achieve this. With Exadata the focus of course was to increase market share and revenue, but do so while fixing a problem. They have almost achieve it.
It does not matter who the vendor is - Disk Drives are slow and investing in management, scale or features for that platform is a waste of Oracle's and the stockholder's money.
The next generation is Flash enabled arrays, Not SSDs in the existing products, actual arrays based on Flash.
Larry knows he needs a flash solution and I suspect he knows that the SUN F5100 is not going be the goto product.
Thing about Datadomain and dedupe in general being the 'new tape' is that DD have not been going long enough to see what it means having to restore large data sets data from 5 years ago, after first rehydrating it, which I think will be no quicker in operation than big restores for a tape system.
Of course your average EMC salesman will be there to support his customer 5 years hence, if they are having a challenge waiting for the array to piece together the critical data.
And of course tape is very cheap to power compared to keeping all your deduped data continually spinning on HDD, in case it is needed. The operational life of HDD arrays is significantly less than for tape, and the product life cycle reflects this too.
And of course EMC will doubtless change direction again in 2 years, just as they did with Quantum. You DataDomian fans just keep paying for the licensing and the power.
I don't think Cesare is much of a Guru. His thinking is very last year. Disc is on downward slope, flash and tape on ascendency. IDC think same for their 2011 storage predictions.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
