Closed seano-vs closed 5 days ago
Right now, this means we can only scale deployments down to 0 and back up to 1 instead of doing a rolling restart
I found this >_< https://github.com/digitalocean/csi-digitalocean/issues/136
also, now the web and sidekiq pod need to be on the same node, so I've hard coded it into the deployments. I'm concerned that this may not survive a mandatory docean k8s cluster upgrade
OOF. This is bad. Hard-coding the node name does not persist across upgrades >_<
This is now going to be high priority to fix. I just applied an update to some of the k8s config and it brought everything down ;___;
Ever since the cloud migration, we've been getting multi-attach errors when scaling out the web service. This is probably due to the RWOnce nature of the volumes. This will need to be resolved to scale.... like, at all