On 26/06/2023 13:48, Fabian Arrotin wrote:
Hi,
Due to small (unscheduled) outage we had with Duffy earlier today (more info on https://pagure.io/centos-infra/issue/1204), we'll have to pause Duffy API to ensure Duffy DB and provisioned aws infra can be reconciled.
Migration is scheduled for """"Tuesday June 27th, 7:00 am UTC time"""". You can convert to local time with $(date -d '2023-06-27 07:00 UTC')
The expected "downtime" is estimated to ~10 minutes , time needed to :
- stop duffy API service
- terminate all Duffy EC2 instances (including the ones in "unknown"
state, from a Duffy DB PoV)
- clean up Duffy DB to reflect "starting from scratch" state
- restart Duffy API service (from that point, Duffy will reprovision all
the pools automatically)
Forgot to reply here earlier (but we were following-up in #centos-ci irc channel) but here we go : maintenance took place and took ~10min and after that Duffy api was back online and provisioning new EC2 pools, now being used