Back
Similar todos
reduce 
#prose server cpu usage by disabling Nomad log collection on services which do not require logging
Hetzners dedicated server blew up on the weekend, but on Monday I concacted their support, and they had to fix the motherboard. Good thing it was on a read mongodb replica
fix broken nvidia a100 gpu server at vultr which has put 
#crisp mirage down for whole night due to being out-of-stock and no replacement physical node could be allocated
wake up from mayhem with my NVIDIA A16s, 70% of them were down all night and only fixed this morning by Vultr. A40 for large LLM still up fortunately #crisp
fixed digital ocean sever, now using dedicated cpu for 
#killspam
discover new issues with server timeouts 
#lyricallabs
fix server issues 
#shakedos
Add swapfile to servers after OOM issues, and come to conclusion that generally servers always should have a little SWAP by default #shipnix
Temporarily break 
#wip when trying to migrate to Digital Ocean, but requests started timing out so I reverted the change
Stress test #shipnix by provisioning 10 servers simultaneously without running out of resources. Success!
🛠 moved the two last crons running on my laptop to a remote server (they were using puppeteer to control a chromium instance) 
#life
spend 1h debugging some k8s complexity-induced issue after a node failure at cloud provider for #crisp mirage gpus
🚀 refactored my server setup, shut off a DO droplet, redeployed two dozen websites, and configs are getting smaller and much easier to support
fixed broken digital ocean server
⬇️ spun down a bunch of Fly.io machines and updated them to only run on demand. 
#servers
realized tonight I can halve some of my server bills by consolidating a few servers and turning off a Redis server #servers
fixing server failure on massive reqs 
#fms