9 comments

[ 4.8 ms ] story [ 29.5 ms ] thread
We did see some http errors and latency spikes on our servers at us-central-1c but it lasted for a minute or so, without any downtime.

Have no idea if other regions or users had major issues though.

This would explain the "phones are down" messages I have been getting. It apparently resolved before I was able to get to a terminal. us-central1-f if anyone is interested.
Well architected applications be like... so what :|
Sounds like a small scale outage, I wouldn't use that headline TBH. Amazon wouldn't even report something like this. We get various small outages from AWS every few weeks, and it doesn't even show up on their dashboards.
> We are still investigating network connectivity issues for a subset of instances that have not automatically recovered. We will provide another status update by 09:00 US/Pacific.

Sometimes they do. If enough customers are affected, Amazon is "required" to disclose. The number, however, has never been disclosed to the public so we don't know the rubric. I have seen announcement about partial AZ failures myself.

Google doesn't say how many customers and how many instances are affected by this, so I will take it as a partial AZ / DC failure (a good number of racks are having issues).

SoftLayer sends out outage alerts for EVERYTHING
Google cloud engine doesn't even exist. Google cloud is a platform that provides a product called Compute Engine. I use it. it isn't down.

a small subset of instances, from a single region are affected.

you're all idiots.