« February 2008 |
Main
| April 2008 »
There has been a lot of attention, good and bad, surrounding the collision of Autodesk applications and Wide Area File Services (WAFS) solutions. (Note: I use the term WAFS because that is what most people are familiar with, a better term to talk about the functionality is Wide Area Data Services or WAN Optimization). Two Autodesk applications, AutoCAD and AutoCAD Civil 3D, create problems for any technology that relies on de-duplication capabilities. Customers configure the 'save' features on certain application files and the applications scramble all the bytes of the file regardless of whether any changes were made. (OK, I am done with technical stuff). The problem is that when customers attempt to share or store these files, they cannot be de-duplicated because of all the byte scrambling which means that there is no way for WAFS solutions or storage de-duplication solutions to impact (or improve) performance and capacity.
To date, most of the discussion has been around WAFS because many Autodesk customers are architects and engineers that share files across remote offices and need the performance boost. It appears that all the WAFS vendors (Riverbed, Silver Peak, Cisco, etc) are impacted and are working with Autodesk to resolve the issue. While this situation may not impact your life, there are some 'read between the lines' impacts that everyone should be aware of and that will be this week's list.
- There is an application out there besides Oracle, Exchange, SharePoint and Web 2.0 YouTube stuff (I really don't know what this is, but everyone talks about it so it must be real).
- Autodesk application developers probably did the 'save and scramble the bytes' development for a reason. They probably had no idea what the downstream impact to WAFS and storage systems that have de-duplication features would be. (De-duplication solutions look for bytes that have not changed or aren't already stored and the Autodesk files would constantly trick them like they trick WAFS.) It proves that no matter how functional the acceleration and de-duplication products are, there needs to be application knowledge.
- There will likely be more applications that collide with what WAFS and de-duplication vendors are trying to do. I think this is a very good thing because the application developers will soon understand that what they do will impact the underlying infrastructure.
- Prior versions of Autodesk applications did not create this issue. It only happened after upgrades to a more recent version. Customers should ask questions about how files are saved, transported and ultimately stored when they are thinking about application upgrades. This line of questioning can help preserve investments in WAFS and de-duplication products.
- It is clear that WAFS solutions like those from Riverbed and Silver Peak are adding significant value for Autodesk customers. (If they weren't adding value, losing the benefit wouldn't be headline news and I would be blogging about college basketball right now). I am sure that these solutions can do the same for other applications and ESG has seen some of them in action via our Lab group. If you want to make the most use of your existing bandwidth, whether it is for application file sharing or for remote replication, WAFS makes a ton of sense (translated - you want to do these things but don't want to buy more network capacity. Therefore the only way to do them is by buying something that makes the network bandwidth you have more efficient).
It is allergy season in California and the Tech Wife is miserable. I am miserable too, but mine could be allergies or a sinus infection. Regardless, neither is fun, making this time of year anti-climatic. It is great that 'winter' (I am using quotes because California, in my opinion, does not really have winter.) is finally over, but it is hard to enjoy the trees and flowers blooming when you are sneezing every fifteen minutes. This whole situation got me thinking about things that I would be allergic to if I were an IT consumer. (It's a fun list this week because I am doped up on Benadryl.)
- 3 year hardware warranties. If you extend them, the vendors make you pay a fortune. If you do not extend them, you are buying more stuff. Sometimes a fourth year just makes life easier.
- 90 day software warranties. Who installs enterprise software in 90 days? You might as well prepay maintenance for three years because the initial warranty is useless.
- Any vendor presentation that shows a financial services customer as a reference. Yes, I want to run my business like Bear Stearns and Countrywide. (I would prefer a vendor that has a customer in the high performance computing industry. Those organizations beat up products really good so if your product makes it through, it must be decent.)
- Buy 2 boxes (servers, storage, etc) and get the third one free. 'Buy one, get one' deals are best left for supermarkets and other retailers. There is something wrong when a vendor starts throwing in extra equipment just to get your attention. I am all for throwing in software, but extra boxes crosses the line.
There are a lot of things that are changing to solve these allergies (Software-as-a-Service, etc), but until vendors hear more sneezing at these tactics, nothing will change.
Let’s not kid ourselves, we are in a recession and bad news is all over the place. Credit risk, Bear Stearns, and other headline stories are making for a morbid beginning to 2008. I am already looking for the bright side of this ugliness and hope that I never hear the word ‘subprime’ again. The current worldwide economy will impact IT budgets and ESG is already receiving inquiries that revolve around customers getting the most for their money. Here are some ideas on solutions that may help you weather the IT recession storm.
Data de-duplication/compression anything. Our research suggests customers are getting between 10 and 20X data reduction from these solutions which means they don’t have to buy as much storage capacity. Wide Area File Services or Wide Area Accelerators. These solutions help maximize existing bandwidth by intelligently identifying the bytes that have already been sent from one site to another and never resending the same bytes twice. There is a much more technical explanation, but I will let the vendors tell you all about it. These solutions make remote office application performance better and facilitate disaster recovery without requiring a huge pipe between two sites. Thin provisioning. You can trick applications and databases into thinking there is more storage allocated than there really is. The best part is tricking the administrators who always ask for more than they need. You can avoid wasting storage. Server virtualization. Enough has been said on this topic. Power and cooling savings and reduced hardware expenditures make this an easy investment to rationalize. Try and use anywhere that is feasible. Search/review/archiving software. These are electronic discovery process staples. If your company does business in the United States, you will be looking through e-mails, files, backup tapes and a host of other data sources. It is only a matter of time. You can either pay attorneys by the hour to read files manually or make an investment to automate some of the process. You may actually be able to use some of this technology for other things which makes it easier to justify.
There are probably more that I cannot think of right now, but this list should get you through the next couple of months. Think positively.
This one is going to be quick because I am on the road again and the time zone (plus the bars) are killing me. I could have lied and said I was really too busy to blog, but what would be the benefit in doing that?
NetApp held its Analyst Day on Tuesday outlining its financial targets for Wall Street and detailing the reasons why customers will look to them for storage and data management solutions. I am at said event thus the traveling excuse. My very quick summation after speaking with NetApp executives, investors and customers over the last 24 hours:
- NetApp unveiled a new brand, logo and company name. Kudos to their team for pulling this off while growing at roughly 20% per year. The 20% number is significant because it proves NetApp didn't get distracted from its products or customers while doing some necessary marketing research. Customers see the new logo and tag line but it is hard to appreciate what this takes from an investment and manpower standpoint. The only thing they see is products that continue to work.
- I met a customer who took 3000+ physical servers to near 200 with VMware and NetApp. After hearing the reasons why the customer made these two vendor selections (versus competitors' offerings), it is hard to argue why these two companies won't have success as partners and as separate entities going forward.
- It appears that NetApp is preparing an executive management transition as Tom Georgens keeps getting more responsibility - this time being named to NetApp's board. Customers will appreciate Tom's candor and operations prowess as it should translate into even better product quality and faster innovation / development cycles.
- The one thing I see NetApp missing is (more) software that enables customers to put additional data into their own storage. EMC has Legato, IBM has Content Manager and FileNet, HP has IAP and Data Protector. NetApp has Symantec and Commvault - two very good 'partners.' But I bet I could find a few folks who would love to get backup and archive software from NetApp. Call me crazy but somehow there are benefits when customers deal with fewer vendors.
I recently participated in Dow Jones VentureOne Summit. I sat on a panel that discussed virtualization trends. The room was full of venture capitalists who asked some very good questions. However, I spent some of my time trying to refute many of the myths around 'virtualization'. Luckily, I had my ESG virtualization research with me so I could support my opinions with real data. This is a quick list, but here are the worst myths out there.
- No one has deployed server virtualization in production environments.
- Microsoft will not have an impact on the server virtualization market for another 18 months.
- Desktop virtualization is not ready for enterprise customers
- Non U.S. markets are really slow to adopt virtualization solutions.
- There is no innovation left for hypervisor vendors. It is all going to be management.
I could spend time refuting these as I did during the panel, or you could read our research or set up a discussion with us. There is no reason why these thoughts should continue to permeate the market.
My middle name begins with an ‘A’ making my initials ‘B.A.B’ and my last name is Babineau. I give my parents an A+ for creativity and actually putting some thought into it (this was not easy for my mother who surely was miserable given the multiple days of labor). What does this have to do with technology? As an analyst, I meet with a lot of companies that have good products, good management teams and the worst (sometimes silly) names. There should be some creativity and some practicality with the name of a company, especially an emerging vendor. Don’t get me wrong, a vendor’s product should speak for itself, but a little bit of name recognition cannot hurt. The ‘I like the Company Name Because it Makes Sense’ list:
Attenex (Pronounced ‘A-Ten-X’) – Attenex was founded on the premise that corporate counsel would see a 10X increase in electronic evidence review productivity. Instead of reading e-mail and Word doc printouts, attorneys can use Attenex to find relevant evidence. I am not sure if everyone actually sees a 10X improvement, but I have spoken with many of their customers, and many rave about the productivity enhancements.
Gear6 – This is a play off of 6th gear which may be equated to performance. Gear6 makes cache systems that sit in front of a storage system, making the device run like a Maserati even though you may have paid for a Camry. The company makes performance a priority and can help customers get the most speed out of their storage. If you need to kick your disks into 6th gear, this isn’t a bad place to start.
Data Direct Networks – This one is a little more ambiguous, but still gets a nod. Getting data directly out of a storage device seems like a menial task, but these guys are doing a good job. This company sells a dense and very fast storage system that is widely used by HPC and media customers. Also, many Web 2.0 giants that developed their own applications that require screaming fast storage buy into Data Direct’s storage system architecture.
Index Engines – An engine that indexes something? Yes. This company has technology that enables customers to process, index and search backup tapes. In case you may have missed it, litigators and regulators are relishing the evidence found in tapes. ESG research suggests that over a third of the electronic evidence asked for by attorneys was originally created and stored 3 years before the request. This means that companies need to know what is on their backup tapes. Somehow, Index Engines has figured out how to provide this insight and if you have ever paid for a legal service provider to do the tape restoration and indexing, you are well aware of how easy it is to rationalize an investment in a tape indexing engine.
The Tech Wife has her own opinion on company names, especially the ones she hears me talking about on the phone. I will refrain from sharing that opinion, but let's just say her common reaction is, “Who came up with that name? I hope they are not in marketing.”’
|
 |
|
