Anil Dash from Six Apart gives an interview post Typepad crash. Speaking as someone who is currently evaluating options for blogging infrastructure right now, I think it's fair to say that the idea that 6A would be able to sell me the Typepad service to host SAP's blogging initiative is kind of whimsical. As much as I like the management interface of MT, especially the ability to manage a bunch of blogs through a single console, and as much as I like the idea of sidestepping IT with a hosted service (which is what we are doing with Socialtext) I really have to say that I would not be comfortable sponsoring any proposal that had Typepad as the provider. It's not just the crash this last Friday, but the entire range of performance issues, up-and-down statistics service, and the general impression of infrastructure fragility. Service providers live or die on their reputation as well as their feature set and economic value proposition, it's going to take 6A a long time to rebuild their reputation as a reliable hosting provider.
The only other comment I have to make about this is that I have seen it posted elseware that this is a failure of web 2.0. What the hell makes Typepad web 2.0, it's a big frickin ASP? Please, somebody tell me because I would like to know.
Link: Exclusive: TypePad outage update and details.
"We have the kind of flagship customers that some people were blogging about today like MSNBC or ABC News or USA Today are all running their blogs on TypePad. That's an incredible responsibility. It's just as serious to me as people who have their baby pictures and talk to their friends and family on there. That's an incredible amount of trust to put in us and I am at least heartened that it made the news that we let them down today. You want it to be newsworthy when you fail, not when you don't."
I think Om Malik et al. are off base in whole debate.
The problems that SixApart are having simply shows that the company lacks key expertise in engineering operations. Nor are SixApart the only culprit. I have yet to see anything that indicates any web company has the expertise for engineering operations. There is nothing inherent in these companies that should cause the failures seen.
All I see is a lack of risk analysis and a lack of proper planning. What makes this disappointing is that the tools and methods for engineering operations are well known. Engineers have been developing them for decades. Why aren't they being used? Is it a general disdain for things outside of the Internet? Or arrogance about the superiority of Internet technologies?
Until web service companies gain the expertise in engineering operations, they are not ready to be mission critical systems for companies or for people.
Posted by: Simon | Dec 19, 2005 at 01:52 AM
A point that I keep making... yes, building large scale datacenters is difficult but it's been done many times before. There is a systems component and a process component, both of which appear to be at the rool of the latest Typepad outage.
Case in point is Salesforce.com, they operate a very large and complex hosted app successfully. You don't hear about massive and widespread failures affecting that company.
Posted by: jeff | Dec 19, 2005 at 07:14 AM
Yes, I agree. The interview with Anil Dash indicates they failed to test a component properly and when it failed they didn't have a process in place to immediately recover.
Given the rest of the interview and the general blogsphere comments, I really wonder if anyone is actually going to learn anything. The whole episode (and year for that matter) is being shrugged off as no one's fault or inevitable. I guess it is going to take a few companies going bankrupt or losing major customers until the issue is really acknowledge.
Posted by: Simon | Dec 19, 2005 at 09:13 AM
I think we spoke to soon Jeff. http://www.infoworld.com/article/05/12/20/HNsalesforceoutage_1.html
Posted by: Simon | Dec 21, 2005 at 08:50 AM