Mutant array upstarts feast on EMC, NetApp's leavings Mainstream storage vendors have a potentially huge blind spot in their product strategy and hybrid array startups are now eating away at their customer base. Nimble Storage, Tegile and Tintri, specifically, appear to be benefiting at the expense of mainstream storage giants EMC and NetApp. When the storage array mainstream met …
EMC Fast cache - which seems to be the most sophisticated and advanced SSD caching of the enterprise companies as it provides both read and write caching. (as a staunch 3PAR supporter it pains me to admit this every time, sigh - though these new upstarts aren't mature enough for my liking - so I'm happy to take the trade off)
Netapp's cache by contrast is a read cache only. (read caches wouldn't do squat for my 92-95% write workload)
Saying it accelerates writes is stretching things a little, it extends the cache by providing a larger write cache buffer. But that buffer is really of more benefit in accelerating reads since all writes still go to DRAM first. Overrun the DRAM cache and you still have a problem destaging.
Migrating data once a day between tiers doesn't constituent as sophisticated. Nate -- I'd be more than happy to walk through the architecture differences of transparent well sized flash as a read cach setups versus rocket + turtle setups that you are trusting for 95% write workloads.
"- Neither EMC nor NetApp can readily retrofit VM-aware storage abstractions to their storage SW. Score one to Tintri."
Both vendors leverage VASA and VAAI and have plugins for management from vCenter and visibility into storage.
The advantage the big vendors have, and I'd argue EMC has a bigger advantage here, is the ability to not only work with VMware (uh hello, shared R&D, shared corporate execs, etc...), but effectively backup, replicate, secure, archive and scale. These hybrid vendors have good products, but niche.
What does this mean?
http://www.tegile.com/blog/george_tintri
"...gave me the platform to clear the air on how Tegile’s wicked smart engineers have used ZFS asa base platform for many of the boring parts of storage (who really wants to write their own NFS stack these days anyways??), and focus their time and energy on maximizing our value and differentiation in the market..."
Confirmed. Tegile is running Solaris and is based on ZFS.
https://www.facebook.com/permalink.php?story_fbid=331644460203544&id=196835830351075
"...Tegile Systems: Hi Stephen - Some of the baseline functionality is on ZFS, but we've put a lot of work into areas we found needed improvement or where we built our differentiation...."
Also:
http://communities.vmware.com/thread/398035
"...The Tegile OS is based on OpenSolaris/ZFS with some custom improvements that Tegile calls MASS (Metadata Accelerated Storage System), which implement deduplication and compression in a different way than ZFS...."
Tegile is just a Solaris / ZFS combo. Just as Nexenta or several other storage solutions.
And more confirmation:
http://www.theregister.co.uk/2012/06/01/tegile_zebi/print.html
7x IOPS? That is good indeed, with ZFS dedupe. Today, Oracle's implementation of ZFS dedupe is not good enough. Oracle should buy Tegile and use their dedupe instead, because Tegile's ZFS dedupe implementation seems to be really good and increases performance whilst using low amounts of RAM and cpu. Tegile seems to know what they are doing. And they are cheap too.
"...Confusing a server filesystem and enterprise storage again..."
What do you mean? These Enterprise storage systems are basically just a server with lots of RAM and CPU, and some flash/SSD based caches, and some SATA disks or SAS disks. They just run some special software. Even EMC and NetApp confirms this, but they also add in hardware raid cards into the mix.
So where is the difference between a enterprise storage system, and these ZFS storage systems? Virtually none. Oracle sells their big 7420 Petabyte servers that beat NetApp in price and performance, and they are exactly as I have detailed: server running special software. In Oracles case it is Solaris/ZFS. Tegile does this too. And both are Enterprise. There are people buyilding similar servers by themselves, running Solaris / ZFS.
Just build a heavy server and run ZFS on it, and you are basically done. (Not quite, but almost like this)
Enterprise storage is much, much more than adding a file system to a white box server, You're a server guy with pretty much zero experience of real enterprise storage (that shows). So I doubt I can convince you of the difference until you suffer a long outage or lose a lot of data to your pet ZFS. If it were that good it would have made a big impact in the commercial space and as it stands it has pretty much zilch footprint on proper storage other than the odd dark archive built on COTS gear..
There is an article here, where EMC and NetApp executives talks about the new kid on the block: Nexenta which offers OpenSolaris/ZFS servers. EMC and NetApp says they are using normal commodity hardware running their own software ontop. Just like Nexenta does. And Tegile. COTS.
So I dont see the difference between ZFS servers and EMC and NetApp. I am just citing NetApp and EMC executives. Do you mean they are wrong? They dont use commodity hardware off the shelf? Maybe their engineers have been lying to the executives?
Or what do you mean?
Yes EMC Netapp etc take advantage of COTS hardware for some parts of the product lines, however they provide much, much more software engineering around Reliability, Availability & Serviceability of the platform as a whole, they also provide end to end qualification between O/S hardware and middleware as well as combined hardware and software support. Those are just the tip of the iceberg, being the most obvious areas where they add value, which is reflected in overall platform cost. Just adding a file system (no matter how advanced) on top of a generic O/S, sitting on COTS hardware does not an enterprise solution make.
Correct, Tegile is a ZFS based storage system. So has all the pitfalls of ZFS - most notably the performance cliff when capacity utilisation climbs (starts at 30% used).
There's an outstanding blog by Delphix which details this exact problem, which is available for viewing here:
http://blog.delphix.com/uday/2013/02/19/78/
Note: Neither Tintri or Nimble Storage is built on ZFS - both have their own purpose built & designed file system.
I must chuckle at all the AC's posting. I wonder how many work for the "mainstream vendors" as Mr Mellor called them...
RE the FUD about EMC working with VASA/VAAI and so would be able to do the same as Tintri - there is no way that any of the Tier 1's would be able to come anywhere near what Tintri have been able to produce in the next 18 months, without a complete ditch and shift of their core underlying OS and file system and thus install base (something called The Innovator's Dilemma - exactly what Netapp hit EMC with in the early-2000's). Much like how EMC nor Netapp would not be able to match the IOPS, capacity, protection & scaling that Nimble Storage can reach using the same core hardware.
PS - I have no problem stating that I work (and love working for) Nimble Storage, as more deals vs the "mainstream vendors" are being won than lost right now.
Wait, wat? NetApp has had efficient primary deduplication for years. What do you mean they can't do it?
As for tintri, either one COULD do it, the "magic" is the fact they only provide nfs access, so they simply monitor the access to files on their filesystems. Whether they find it valuable is another question entirely. The overhead of that monitoring would likely break both for general purpose workloads which provides them far more marketshare than tintri could ever hope to capture.
It depends what you consider efficient. NetApp can do about 2x deduplication on a good day, which about pays for the overhead of WAFL and RAID-DP. It also only runs in the background, so there's no in-line dedupe, which means that there's a buildup of redundant data followed by a slow period when the maintenance is being performed.
@ATJ
Strange that you didn't respond to the so called FUD and just suggested it was all addressed in a whitepaper somewhere. From experience what Tom said about Netapp dedupe is correct, if it really were that good, you'd charge for it like you do everything else. No such thing as a free lunch.
"Neither EMC nor NetApp can retrofit efficient primary data deduplication to their legacy storage. Score one to Tegile."
As AC above, yes NetApp did "retrofit" efficient primary data deduplication to its Data ONTAP storage operating system with Data ONTAP release 7.2 in 2007,
Also, because NetApp deduplication leverages block checksums which were implemented into Data ONTAP's WAFL filesystem long ago, means there was no "retrofit" required. All that was required was to leverage the existing checksums with a deduplication software module.
It's tough to convince me that these upstarts are feasting on the established players when they have <200 customers (last I heard there were over 50,000 VNX systems sold). I will give credit to Nimble though with 2,000 boxes in the field. The product is hot right now in the Midwest US area from what I'm seeing as a consultant in the field. While I don't understand their secret sauce yet, they seem to have figured out how to really make an SSD cache work in front of all SATA, while EMC/NetApp talk about doing that in marketing slides but real world configs never leave the factory without a healthy amount of 10K/15K.
HSG -- I know Netapp has an offering similar with their flash cache offering. The problem is it is only available for their midrange and up. That and their boxes suffer from the efficiency tax that is their file system/data layout. It is no doubt elegant in design and still peerless, but it has not been designed with the low end in mind.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
