Hey, as most of you know, IBM purchased Storwize and changed the name. I have updated all of my previous videos that talked about the technology from saying "Storwize" to "Real-time Compression" - and added some more technical details. Also, with the help of Media Boss, I have updated the intro - pretty cool stuff! Have a look and tell me what you think!
Over the past few weeks I have witnessed the proverbial mudslinging that takes place in the blogosphere when marketing feathers are ruffled. Most recently I was reading Rich Anderson of The StorageSavvy Blog. The article was "Compression better than Dedup? NetApp Confirms!"
I have to agree with Rich on many fronts. First, "When all you have is a hammer, everything is a nail." Rich points out vendors have to sell "what's in the bag" so it is conceivable that all problems look like they can be solved with their solution. If you look back over the last few years NTAP has always had a "me too" reputation. Whatever the industry has, they have one too and its better. For the last few years, while competing against Storwize, they have pulled the EMC tactic of trying to stall a market by saying, "We have optimization for primary storage with deduplication." The reality is, you can't use it in real time, it is a resource hog, and again Rich mentions, the only use case it works well on in primary storage is VMware (and that is ONLY IF the customer stores their data outside the .vmdk file otherwise compression is much better). Now that NTAP has compression their story has changed saying that compression on primary storage is better for most use cases. Duh! The folks at Storwize (now IBM Real-time Compression) have been saying that for years. Why, deduplication is great for repetitive data sets, i.e. backup, not primary storage. There just isn't that much repetitive data in primary storage. Again, NTAP is trying to stall the market saying they have "in-line" compression for primary storage. Sorry guys, not good enough. In-line is NOT Real-time. Rich also points out that the key characteristics of storage for customers are capacity and performance. Patrick Rogers of NTAP has said publically that compression WILL indeed impact performance and that they even have a tool that will tell you how much performance will be impacted. While NTAP may say compression is "free", we all know nothing worth having in life is free, you get what you pay for. If you need the performance to do compression you are going to have to perform a major upgrade to your filer in order to just be able to perform compression let alone try to do compression in real time. No real savings there.
Today IBM is making one of the most significant storage announcements in the last 10 years. (Steve Duplessie is quoted as saying “…last 20 years.” but announcing that they were basically getting out of the storage business in 2000 by selling their Mylex division was pretty significant.)
Today IBM has made it abundantly clear that they are back in the storage business. It makes a lot of sense actually. Servers have been commoditized to about 3% margins and services is a body business (to make more money you need more bodies). Storage is the only place these days to add significant ‘value’ to the infrastructure eco-system.
Storage software – or as I like to refer to them, storage services allow vendors to add more value to commodity hardware by providing very useful capabilities such as thin provisioning, virtualization, snapshots, replication, and optimization solutions such as real-time compression.
However this story is as much about the naming of the product as it is the product itself.
Let’s rewind a bit. It is September 1st in Tel-Aviv. IBM has just completed the Storwize acquisition and we are having our leadership meetings discussing the integration when the new management team informs us that: “You know, we really like the Storwize name a lot.” “We like the name so much that we decided we want to use it for a totally separate and new product.”
As new employees we two choices:
Having branded products and technology before, I didn’t really see this as an issue. Sales of course didn’t care for it. My feeling was, that even though the new Storwize V7000 doesn’t support the new IBM Real-time Compression (which now you have it, our new name – not flashy but does describe exactly what we do), it is the most modular storage architecture designed today and I am quite sure, that because the value in this ‘platform’ is really about storage services (software), IBM Real-time Compression could eventually make it on to this platform. That being said, the amount of noise IBM is going to make regarding the new Storwize will drive the old Storwize sales team to a number of new opportunities. Sure, we will have to work harder to vet them as the new V7000 is a block device and IBM Real-time Compression is an appliance for NAS today but that is okay. We will now have exposure to a great deal of customers to tell our story.
Storage is elastic. How do I know you ask? Yesterday I visited a customer who is using the Storwize product to do Real-time Compression on their primary storage. The customer is Allianz and has been using the product for over a year. They see 75% compression on their users home directory data. To give you an idea, Allianz is an insurance company and generates TONS of spreadsheets, 14TB worth of spreadsheets (okay, not all 14TB is spreadsheets but you get the picture).
Prior to Allianz purchasing the Storwize technology, Allianz didn’t have great data management practices. Users store data in their home directories and there is really no discipline around deleting or cleaning up files so data just grows. Additionally, storage isn’t really budgeted for. Overall IT is but at a storage level, they just purchase some when the need some.
Again, prior to the Storwize technology, Allianz had their primary storage and a backup to tape at their local site. They then replicated the data to their remote site and also performed a backup to tape.
Allianz has an overall IT mission to reduce spend by 10% per year. The thing to think about is that this 10% could come from a lot of places including data management.
Once the Storwize technology was installed the first things they saw were:
The new IDC report that came out in May stated that the "...amount of information in the digital universe would grow by a factor of 44 and the number of containers or files would grow by a factor of 67 from 2009 to 2020, the number of IT professionals in the world will only grow by a factor of 1.4". There can't be a more clear indication that it is time to get your storage under control.
Video created in conjunction with !
15 Minutes Can Save You 50% or More of Your Storage Capacity! Learn How with Storwize!
Click here to learn how Storwize can help you get your storage back under control!
I am not sure how many of you have seen the ABC show 'V' but it looks like they have it looks like they have two things in common with Storwize.
First, if you are unfamiliar with the show V - the description from their website is as follows:
The Visitors spread their message quickly and clearly: They come in peace, bearing gifts of medical miracles and technological breakthroughs far beyond our wildest imaginations.
Very similar to Storwize. We visit customers in peace bearing gifts of miracles and technology breakthroughs beyond their wildest imaginations. Storwize is the only company to do real-time data storage compression without performance degradation. As the growth of data exceeding 487M TB (IDC 2008 study) something has to be done to control this growth and end users are quickly turning to capacity optimization technologies to control this growth and more specifically to Storwize Real-time compression.
With another banner quarter under the belt, I had an opportunity to meet with a number of existing customers as well as customers to be. One of the more interesting meetings came when I met with a large manufacture in the mid-west. We were sitting with the storage architecture team and the finance team. We were going through our usual presentation, showing the finance folks the savings they were able to achieve with Storwize. Storwize was able to compress their data 66% giving them back more than half of their storage capacity that they could continue to grow into. We also showed them how there was no impact to their production environment with the Storwize appliances in line. The finance folks were VERY impressed. At that moment one of the storage architects jumped out of his seat and said, "But wait, your forgetting the most important point." To which we all looked at each other a bit perplexed and said, "Oh, what is that?" And he replied, "Storwize is Butt Simple!". He went on to say that the deployment (during their POC) caused no down time (they were in an HA environment) and once in place, they just started compressing - no change to their filers, no change to their networks, no change to their applications and the end users never even knew anything different. He said, "Storwize makes my job simple and look like a hero to my boss!".
Rather than leave a lengthy comment on (and from a marketing perspective, Friday’s are bad days to post blogs – especially in the summer) – I thought I would respond here (this may get lengthy as Tom made a number of points which I need comment on).
The first thing I do want to say is that when doing technical marketing; the proper strategy would be to not be on defense but rather take an offensive approach. However, given the amount of FUD that Tom put in his latest blog post, I have to defend compression to some degree.
Now, I think we can all agree that data compression and data deduplication are two technologies that can complement one another very well. Avamar (EMC) deduplicates the data at the source and then compresses the data before sending it to the Avamar Data Store gaining tremendous efficiency in network utilization. ProtecTIER (IBM) compresses the data once it is deduplicated at the target device before it stores the data. Other solutions also combine compression and data deduplication.
I’d like to comment on some key point Tom made in his piece where he is just blatantly wrong:
1) Compression identifies redundant data across a very small window, usually 64 KB. – While this may be true for other compression technologies, this is not true for Storwize. Storwize performs compression where the initial window is not fixed in size at all; it is the resultant write that is fixed in size. This size is also specifically mapped to the I/O patter of the data being written. The goal is such that in 1 I/O Storwize can do all the work it needs to on a particular file or LUN and it is for this reason Storwize has no performance penalty.
2) Compression produces data reduction rates at most 2X for most data types. – Seems Tom needs a lesson in the most common answer in IT – “IT DEPENDS”. Data compression ratios are 100% tied to the data type. For a true indication of data compression ratios see Figure 1.
There was an interesting regarding Permabit who is now providing primary storage optimization through OEMs and having their solution embedded into the storage system. This further drives home the point of . I do have a couple of questions however:
1) What is the performance like? I see phrases such as “High Performance Data Optimization Software” but don’t see any performance metrics – such as ‘no performance degradation’ for customers utilizing the solution. Or testing metrics from their ‘partners’ (as it probably isn’t in production yet) – which brings up another question:
2) Why were none of the ‘design win’ partners quoted in this announcement?
3) Rehydration – Mr. Floyd states:
Permabit's Floyd claims Albireo can maintain data integrity because data written to disk isn't altered, and the reduction takes place out of the data path. When parallel processing is used, deduped data doesn't have to be rehydrated when it's accessed.
The question is – if it doesn’t need to be rehydrated, then how does the application read it? I can only assume that Mr. Floyd means the data doesn’t have to be rehydrated on disk, which is fine, the question become: a) how does the application know what the data is? (Ocarina uses an agent to help them understand the data, but this is another thing to manage) and b) What is the performance of the system looking up all of the hash keys to reassemble the data on the fly, so how much more storage resources will this consume?
4) Back to performance – Permabit states:
When done inline, data will flow to the Albireo library before going to disk. will write data to disk first, then scan and eliminate duplicated data. The parallel option sends data to disk while still in memory, and applies updates the same way as post-processing without having to read data off disk. Each method has different amounts of latency and reduction efficiencies.