Posts by cult@pony.social
 (DIR) Post #B5FKMkva7g1xm2tim8 by cult@pony.social
       0 likes, 0 repeats
       
       @thephd <3 sorry for that downtime yesterday
       
 (DIR) Post #B5FKYxH034MhXOjV7Q by cult@pony.social
       0 likes, 0 repeats
       
       as a sort of early post-mortem of yesterday's tls expiration: I failed to figure out I currently have no monitoring for TLS expiration in place, that used to be on my DO account but I've removed it with a plan to replace that never materialized. So when the cert began ticking down the expiration I didn't exactly notice. Yesterday i was fighting the cert-manager for kubernetes quite a bit to get it issue a new cert, it seems my load-balancer is taking issue with the setup and not quite happy doing HTTP based cert validation. I'll have to setup some more robust monitoring tbh, and do a permanent fix for the cert-manager that isn't a stupid hack.
       
 (DIR) Post #B5Mg1U1vJLUYNpqSYa by cult@pony.social
       0 likes, 0 repeats
       
       vertically spinning pony
       
 (DIR) Post #B5r7K0kEwXt4Z0Qlzk by cult@pony.social
       0 likes, 0 repeats
       
       gonna be rolling out some patches to the infra of pony.social. I'm trying to finish that with 0 downtime, but things might slow down a bit as data gets shovelled around
       
 (DIR) Post #B5rg3s7CkznHYtDD9M by cult@pony.social
       0 likes, 0 repeats
       
       alright, pony.social should be coming alive again. The update went great except where valkey shat itself and i had to restore it from backup. Was extra amusing because the redis host is just plain unrelated to the rest of pony.social? From the logs it's mostly related to networking not handling traffic well as the updated host wired itself up.