Why didn’t Typepad blog during the outage?

UPDATE: Interview with Typepad’s Anil Dash here

As I’ve said several times in the past on my blog, scalability with Web 2.0 applications matters. The last couple of days of Typepad system failure has made me question whether or not I can continue blogging with this service. At least using Movable Type or WordPress on my own hosted site (with hoster Blue Host) would allow me to completely backup my posts vs. relying upon the blogging host, in this case Typepad.

The effort, energy and investment of enthusiasm in nearly every post makes the continuation of my posts something I care about! When they announced the failure on December 16th, they "rolled back" to their last backup….several days earlier. Did this mean they don’t perform daily backups? If so, they’ve violated "Information Technology 101" rules. As of this writing, images are missing from the last few days of posts.

For many months I’ve been experiencing ongoing slowness and performance issues with the Typepad service. Ironically, I’d emailed SixApart (Typepad’s parent company) CEO, Barak Berkowitz, November 24th from a hotel on my Thanksgiving holiday articulating my experience with this unacceptable performance. He stated in part,

While I understand that you feel that this has to be a TypePad problem I believe it is not.  We constantly measure actual performance from all over the world and as well as at the datacenter on our servers and network devices.  For the last 10 days performance has been very good and it is excellent today.  I was actually checking performance around the time you sent this and it was first rate.

Now I understand that when we have let you down a number of times and even pointed out our flaws it becomes easy to assume it is us again, it is only natural that it will take some time for us to earn your trust back.   But I believe that we will.  Please realize that there are very many things that could cause an experience like yours including temporary access problems like you describe, particularly through a hotel network which use complex gateways.

Besides feeling a bit chastised by Mr. Berkowitz, it’s clear to me now that either the measurement of Typepad performance was flawed or that my experience was a precursor to Typepad’s future failure which has occurred. I also submit that I’m a bit taken aback since I’ve been at an executive level at two enterprise software companies, have used a myriad of enterprise-class web applications, and know when it’s my issue or a network issue. It’s now very apparent that it’s not my issue. I would hope in the future Typepad listens more intently to their customers.

Oh yeah…I tried emailing Mr. Berkowitz again this evening but, "The following addresses had permanent fatal errors –<Barakb@sixapart.com>". Guess their email servers are down too.