Boarding - HA PG feedback / troubleshooting
efstajas
PROOP

2 months ago

Hey @Paulo, as requested, creating a thread here

just created a new cluster:

  • cluster id: a82dbe37-29b7-4797-9e90-13d070d38158
  • env id: abe30bda-19ea-4d24-a9ea-527ff56fbbb1
  • project id: 56cafcfa-394c-46c9-a811-dc3207bad3dc

This one actually ended up a bit more broken than my previous attempt. The etcd instances seem to be failing to coordinate and show as "not in cluster" in the UI. No DB has been elected as primary yet. The same 503 on all HAProxy ports problem is there again too.

Lmk if you want me to try anything 🙌

8 Replies

2 months ago

Thank you! WIll look into it, seems like it's just HA proxy having issues


2 months ago

seems like etcd recovered


efstajas
PROOP

2 months ago

ah yes!


efstajas
PROOP

2 months ago

and the DBs are all in sync now too


efstajas
PROOP

2 months ago

looks like the UI just didn't update on its own, after a refresh it looks good


efstajas
PROOP

2 months ago

so yeah just the same HAProxy problem as before remaining


2 months ago

Ok it should be working now, thank you for reporting it! It was indeed the ipv4 issue, the health check service was only listening to ipv4


2 months ago

Will keep working on the cluster downgrade thing, but I know what's up and can reproduce


Welcome!

Sign in to your Railway account to join the conversation.

Loading...