I’m seating here in the airport lounge waiting for my flight back home and trying to reorganize my thoughts about this last HP Discover.
Well… no news: no bombastic announcements here although there are many interesting developments to mull over.
I’ll pass on commenting about “converged systems” (I do not like primary vendors’ approach in this field) but, if you are interested, you can follow the link and find a very good deep dive article from Chris Wahl.
3PAR
The 7450 model, a product which I’ve already talked about, has received an important firmware upgrade and it’s now able to reach 900K IOPS limit (4K, 100% reads). At the same time HP has adopted a new kind of SSD drive that makes it much cheaper. The good thing here is that, as one should always expect (!), the firmware is also available for existing customers.
According to HP (and IDC), 3PAR growth was massive last year and has recently become the second mid-range SAN in the market… that’s a great step further, isn’t it?
If you’d like to know more about 3PAR, don’t hesitate to download my “Down To Earth Report” here.
StoreOnce
StoreOnce probably deserves a lengthy article all for itself, but I can easily summarize the announcement as follows: faster, bigger and more efficient. In fact the new model, when compared to the previous one, is impressive under many aspects. HP is pushing this product heavily and technical characteristics are remarkable. Looking at the whole line-up I especially like the VSA, although not unique in the market. It allows StoreOnce to be deployed when budget and human resources are very limited (like ROBOs for example).
StoreAll
StoreAll isn’t a proper object store (even if it can now be accessed via Openstack’s Swift APIs) and neither a File Server (even if it can be easily used via SMB3 and NFS). You can access it via all the mentioned protocols (and a few others) but its real advantage is in the NoSQL engine that makes the magic. All the files that have been copied into the systems become searchable (not only metadata but also files content) at an astonishing speed. It’s not uncommon to have results in seconds for searches on hundreds of millions of files! And the product does much more than that… I’m waiting for some more documentation and then I’ll go in depth about it in a future blog post. There are very interesting use cases to be explored. Stay tuned!
HP and Openstack
I must admit I’ve been so very impressed by HP faith in Openstack (both on public and private/hybrid clouds). They are also committed in the development of CloudOS (an OpenStack distribution). As already mentioned admirable but all my doubts remain intact (not only about HP commitment, obviously). I believe the number of end users that will be able to move to this technology/platform is very limited at the moment… but only time will tell.
StoreVirtual VSA
Last Monday I released a new publication which I worked on for quite some time and it’s about the HP StoreVirtual VSA. If you are interested on how to deploy a “hyper-converged infrastructure for the SMB” that’s the paper you are looking for. 😉
Here is the link for the download!
Why it matters
Eventually, HP Discover demonstrated its value this time round as well, especially for HP partners and customers… and as I wrote in a tweet “not only for technology but above all for the kind of people you can meet here”
I do want to come back to StoreAll. When we first heard about it 12 months ago at Discover Frankfurt it seemed already interesting. However I have 2 remarks here: as you said, the true value of StoreAll is not the storage part, in fact if you speak to the StoreAll team they don;t have a real storage background, but the processing value is. If that is the true core value, why shipping it in a physical product in the Storage department? Isn;t this something that should be a combination of Moonshot and JBOD or so? Somehow I am missing the physical added value and it seems a limitation in design.
Secondly to your “it’s not a real NAS product” is very true as well. Although it does use IBRIX for front-end NAS capabilities, it doesn’t bring a competitive NAS product to the market. So my question is: does HP still have a competitive NAS product? Other than Windows Server off course. Maybe related: do they have to? Is HP missing a core storage product here or not?
Ok, maybe a last question to “it’s not a real OBJECT product” as well. Same question; everyone is looking at object storage as we see the limitations of block and file somewhere in the future specifically for new types of workloads/applications. Does HP need to address this or do they have an upramp portfolio we (I) are missing here?
Hans,
thank you for the comment. I couldn’t agree more!
The products has a lot of potential but there are many aspects that should be investigated very deeply.
I hope to receive more documentation soon to dig seriously into it.
I would also like to see a roadmap. 😉
Ciao.
ES