Hacker News new | ask | show | jobs
by mej10 3533 days ago
Even though their status page says everything is basically fine -- oh just a blip with API dns this morning no biggie -- they have had major problems today too. Like, can't connect to dynamodb, sqs, kinesis.

Unfortunately they lie through their status page pretty regularly. I am not sure what it would take for them to admit they had a major problem.
1 comments
shuumai 3533 days ago
Apparently us-east-1 was the only region that exclusively used dyndns (all other regions have failover strategies to other providers and that's why there wasn't downtime anywhere else but us-east and eu-west). They resolved the problem pretty quickly, but it drove our NOC insane...fortunately downtime for us was only about half an hour or so.
link
mej10 3533 days ago
Yeah, thats about a half hour of down time for every AWS East based company... who knows how much lost revenue -- and yet they just say the service is green with a little info tag.

If our service is down for any customer (much less an entire segment of customers) we at least publish that we had a partial outage.

link