I am doing my weekly recap blog a little early this week due to the upcoming Christmas weekend and wanted to share some final thoughts on what I considered the last of the three major storage trends of 2009. Over the last month I have posted a couple of blogs discussing what I considered the top two storage trends of 2009 and this week I wanted to turn my attention to what I consider third major storage trend of 2009: solid state drives (SSDs).
SSDs are anything but a new technology and, ever since I have been involved with storage, the topic of SSDs has come up in passing from time to time. But I guess my first real hint at how disruptive SSDs were going to become came when I was attending C-Drive, Compellant’s annual user conference, in the spring of 2008.
While sitting in one of the one of the sessions, one of the speakers, a Compellent user, was lamenting how difficult it was to optimize a storage array for database performance and how a Compellent storage system had helped him in that area by striping data across the array. One of the Compellant presentors then asked the user if he would have considered using SSD to improve performance and he responded absolutely which I think caught the presenter offguard.
The presenter then turned the audience of end-users and asked them how many would consider implementing SSD technology in a Compellent storage system. Many of them (well over 50%) raised their hand which again appeared to catch the presenter by surprise as he was looking around at other Compellent executives to make sure they were getting this feedback. It was at that point that I knew SSD technology was on the cusp of become the next big thing in storage and it was not long after this conference that Compellent announced support for SSD.
Now here we are approaching 2010 and SSD is on everyone’s mind. Already it is being hailed as the end to high performance disk drives and forecast by some as an eventual replacement for all disk drives – possibly as soon as the end of this coming decade.
Whether or not that comes to pass remains to be seen. SSD first has a ways to go in terms of maturing and in coming down in price to the point where most businesses can justify and afford its deployment. But in this intervening time between its early adoption phase and the end of hard disk drives (HDDs) as we know them, there are three notable ways that SSDs were positioned for use in the enterprise during the course of 2009 that contributed to making it my third major trend of 2009.
First, most SSDs are being initially being deployed on external networked storage systems. Already providers like Compellent, Dot Hill, EMC, HDS, Pillar Data Systems and others have added support for SSD to their own storage systems or can virtualize external SSD storage systems (such as NetApp’s support for Texas Memory System’s RamSan 500) for pseudo-SSD support.
This was certainly one of the fastest and easiest ways to bring SSD to market and implement it in 2009. By using SSD drives from providers like STEC or Pliant Technologies that look like HDDs and then installing them in existing storage systems, the approach only makes sense. However I am already questioning how long this trend of putting SSDs into storage systems lasts before it exhausts itself as companies figure out they are not getting the full range of benefits that they anticipate.
This scenario may take two or three years to play out since more storage providers plan to announce support for SSDs in their storage systems in 2009. However what I eventually see emerging as the problem with putting SSDs into storage systems is the storage network itself.
The current networked storage infrastructures in use in most organizations are Ethernet, FC or some mix of both. Unfortunately neither one of these technologies – even if using the latest and greatest versions of each – currently offer sufficient bandwidth to deliver on SSD’s potential.
The one technology that seems best positioned to break this bottleneck is Infiniband. The problem is no one outside of the high computing space seems ready to commit to this protocol or take a serious look at it. However this forthcoming network bandwidth logjam is already recognized within storage circles and, in talking with CTO’s of a number of storage system providers, some are already exploring the testing of Infiniband within their storage systems in the coming year.
Second, SSDs are showing up in servers. In a trend that largely runs counter to the last decade, storage is moving back inside servers again. This trend is being largely (and I would say almost exclusively) driven by Fusion-io. While I have written a number of blogs about Fusion-io’s ioDrives in the past and the disruptive nature that they pose (Fusion-io puts SSD on a PCI Express card that is placed inside of a server to achieve higher performance than what a high end storage system can achieve), that is only part of the story.
The rest of the story has to do with the elimination of cost and complexity. Many storage networks are still neither easy nor intuitive to manage, especially in high end storage environments. (I do consider 3PAR a notable exception to this rule based upon numerous conversations I have had with its end-users.).
That said, Fusion-io eliminates the need and associated expense to deploy any storage network at all. While its ioDrive reintroduces the inability to easily share unused, excess storage capacity with other servers, its design is so compelling (especially when viewed in the context of cloud computing and emerging cloud storage architectures) that the ioDrive will be a force to reckoned with in the coming years – maybe more so than any other competing SSD architecture.
The third architecture is moving SSDs into the storage network. This SSD architecture I came across this year while attending this fall’s SNW in Phoenix which is currently being championed in the form of the Dataram XcelaSAN. What made this approach so different from the other SSD architectures was that it neither attempts to go inside the server nor inside the storage system but resides in the storage network between the servers and back end storage.
The argument that Dataram makes for this implementation of SSD is that approximately 5% of data on a storage system is active at any one time. However most storage systems only have a fraction of that much cache on their system. For instance, a storage system with 20 TB of active data should in theory shold have about 1 TB of cache that sits in front of it to provide the read and write performance that high end applications need.
A quick check will reveal that many midrange storage systems are lucky to offer a fraction of that with most supporting in the range of 3 – 5 GB of cache on a single system. Dataram argues that by putting its XcelasSAN in front of an existing midrange storage system, it can act as an alternative to storage system cache while turning a midrange storage system into a high end solution at a fraction of the cost of a high end storage system.
Not a bad idea but there are some definite hurdles this architecture will need to overcome. First, Dataram needs to convince people that placing its device inline between the server and backend storage is a good idea. While Dataram has had success and
experience in this space with prior offe
rings, now it is Dataram and not their resellers that must convince end users this is a good idea. Speaking from past experience, this is easier said than done.
Another obstacle it faces is from midrange storage system providers themselves. Depending on how Dataram positions itself, why should companies like Compellent, Pillar and other midrange providers agree to play second fiddle to Dataram and sit behind Dataram’s solution, especially if they are offering SSD within their own systems? Further, since SSD is becoming so lucrative, that means they are giving up the biggest margins to Dataram.
Maybe the third and biggest obstacle Dataram faces is from traditional network storage providers like Brocade and Cisco. These guys never take kindly to anyone intruding on what they view as their turf so if Dataram garners any momentum at all, expect one or both these heavyweights to step in and try to snuff it out. So while I like both Dataram’s architecture and argument as to why deploying SSD in the network makes sense, convincing the channel to sell it, end-users to buy it, existing midrange storage system providers to endorse it and storage networking providers to leave them alone is going to be an uphill challenge.
All this said, SSD gained some solid momentum and mind share among end-users in 2009 and is poised to emerge as a major trend again in 2010. However which of these architectures becomes the predominant one remains to be seen. In the near term (2010 – 2011) I am putting my bet on storage system solutions but as technologies like those from Fusion-io reach critical mass and others from Dataram are tested in the market, the real impact of SSD is yet to come.
Have a Merry Christmas and I will talk to you again next week when I reflect on the most read storage blogs on DCIG’s website in 2009.