slicehost followup

[mark]

Just a followup on yesterday's downtime:

I just spoke with Matt of Slicehost (not sure what his official title is, but he started/runs the place). He had looked into the various incidents we've had over the past few weeks, and it seems that they are all similar: hardware failure of some sort or another.

The good news is, this isn't entirely unreasonable. We made the choice to move to Slicehost's brand new data center in the Dallas/Ft. Worth area. One of the side effects of having new machines is that you have to work out the kinks in them. While it's unfortunate we've hit so many, they are not too concerned about the number, given how many slices we have.

Slicehost is working on rolling out stronger monitoring so they can try to predict failures and better account for them when they do happen. It's sad to have to deal with them, but our system is pretty redundant. There are only two machines (out of 25-30) that take the site down when they fail, so it's pretty bad luck that we've had both of those go down in the past few weeks.

While there is not much action that can be taken at this point, it was a good conversation, and it does serve to emphasize just how top notch Slicehost's customer service is. They are fast, efficient, and take care of problems quickly.

If you have any questions, please let me know!