Hey there, I was wondering statistically what is the most stable region to run on in all of the US and the west coast? There was another incident with SJC, and I want to choose the most stable regions I can so I don’t get calls in the middle of the night.
Thank you
Hi Chris,
The pragmatic answer to your question is that the most stable region to run on is two of them. SJC & LAX; LAX & SEA; any two, but two (or more). But I don’t like having to give that answer and I’ll explain why.
We don’t like to have outages at any level. But on the level of the single host, all hardware fails eventually, and so to prevent downtime we heavily encourage users to run an app with multiple machines and multiple volumes. The failure of a single host server is not a failure of the platform; that is within the bounds of normal operation.
On the other hand, a regional-level outage, whether whole or partial, such as happened last night, is not within the bounds of normal operation. That is our failure to run the platform to the standards we set for ourselves. Every time this happens, it is an internal incident, and every time it happens, we review why it happened and how we can prevent it from happening again. This includes these scheduled emergency maintenances. We will follow up with our providers to learn why the maintenance was necessary and how things can be changed so that future maintenance will not result in regional downtime. There is no regional downtime that is acceptable to us.
You should be able to run two or more Machines in any single region without getting calls in the middle of the night. If you cannot, that’s on us. But since we still have some work to do in this regard, you probably want to run Machines in two regions. Regions are independent of one another, and that should give you the stability you’re looking for.
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.
