For the past few hours we've been seeing lots of H10 errors on Heroku apps in the EU common runtime. Apps crashing randomly (every few minutes), seemingly no reason for it. Anyone else on @heroku seeing this? Nothing on @herokustatus / @asksalesforce yet
@asksalesforce@jamesmgillard@herokustatus@heroku Thank you @asksalesforce - as far as I can see this is a very similar incident to yesterday's (incident 2868 - https://t.co/Bz0DpQMEtF) we are seeing the same behaviour across all of our applications in EU common runtime. Would be great to see something on Heroku status page
Seeing lots of timeout (no headers received) all of a sudden across various Heroku apps. Definitely something going wrong since 18:09 BST but no acknowledgement on @herokustatus
@jamesmgillard@herokustatus Good to know it's not just me! Apps are coming back online quickly after restarting all dynos, but dying again shortly after.
Seeing a lot of instability on our @heroku apps again today, for past 1-2 hours. No incident published the the @herokustatus page though. Is it just me? Seeing outages across about 5-6 apps, no obvious error messages.
@asksalesforce The latest update now states it explicitly:
> Impact is minor and limited to the U.S. region. We are still investigating the root cause.
If that's the case, there must be a separate incident going on which has not been published, because our apps are in the EU region!
@asksalesforce Thanks for that. The @herokustatus update (https://t.co/YASwnCWf3g) refers to US customers only. All of our affected apps are in the Europe (common) region...
@asksalesforce@asksalesforce There is obviously an issue but there is no information on the @herokustatus page at https://t.co/Q0wJ82Yjgq - please use your own status page to communicate with customers!
Seeing some issues with multiple @heroku databases on the Essential tier over the past 30-40 mins. No incidents posted on @herokustatus - anyone else or just me? Connection error: out of memory
@martinbean@heroku@herokustatus Oof... just noticed a small change on our side. The pg:info CLI command was previously returning spurious info (e.g. Data Size: unknown usage / 10 GB). It's now returning the true data size of the database. So hopefully this is an indicator someone is working on it.