30 Hours Downtime

march 7, 2005

It's still going on, yes, I mean the Sidekick service downtime. I wanted to post an updated about it, at least from what I got from T-Mobile Customer Care. I called yesterday afternoon about the downtime and the friendly rep could only give me 50 bonus minutes because I got a $5 credit for the last time the service went down. Well, I just called again (it's 2:30pm PST) and talked with another rep, Teresa, who was really friendly and helpful. I told her that I had been without service for over 24 hours now and I want a credit. So, I got my $5 credit (which if you think about it, covers a week of Sidekick service). It still doesn't help my annoyance that the service is gone, but at least it covers for the downtime. Teresa says that there is no ETA on when the service will be back up. Danger engineers are working on the issue currently though. And she specifically said that the problem is with the Danger Sidekick servers and not with T-Mobile servers. On the HipTop Forums there are reports that Danger Customer Service is saying that everything is fine, but that is not true since I am down and the reports of down Sidekicks has ballooned to almost 600. I hope that the Danger engineers (along with the T-Mobile engineers who are there helping them) will get the service back up as quickly as possible. I just read a posting at the HipTop Forums (which run on a Danger server and is moderated by Danger employees) that this is a known issue and that we should not call our service providers. I say that if you have not gotten your credit for lost service, you should give your service provider a call and get your credit. You are paying for service and if you are not getting it, you should be refunded money for time lost. I have a feeling that Danger is asking that we do not call our service providers because they (Danger) will be getting hit with a loss (at $5 a call) for this downtime. Danger needs to put in some stopgaps to keep this type of service interruption from happening again. A bit of planning for high availability is necessary for any company that provides services for thousands of users. There is no excuse for this type of outage when you are taking money for services provided. Right now, with four days of outages total this year (2.5 days for the last time during the server upgrades and 1.5 days for this currently outage), Danger is running to have an annual uptime of about 98.9% if nothing else goes wrong this year (or if this outage stops sometime tonight). Most companies aim for five 9s of uptime (meaning 99.999%), which means that their services does not go down for any more than 5.25 minutes in a year. Danger needs to start aiming to have their servers up a lot more in order to have better customer satisfaction, even three 9s (8.76 hours a year downtime) would be good. The Sidekick II is the best piece of hardware that has ever come out, but it is strapped to an unreliable backend. Which is too bad.