renstur@coopnet:~/blog$ ls -lt ./topics/field-notes/
field-notes/
-
I edited a Portainer stack's compose file on disk and it stuck
I thought Portainer would discard compose edits made on disk, but on my install the editor just showed them. Stack files are plain files in Portainer's data volume, one folder per stack id. To apply an edit, recreate with docker compose -p set to the stack name, or you get a second project.
-
The Checkmk REST API turned off TLS checks on my alert webhook
I used the Checkmk REST API to keep TLS verification on for my alert webhook. It pointed the rule at a new parameter set with ignore_ssl in it, and the plugin skips verification whenever that key exists. I fixed it in the .mk files and now grep for ignore_ssl after every scripted change.
-
My solar gateway went offline every night at 3 AM, and it was a Wi-Fi channel
My Enphase Envoy dropped off Home Assistant around 03:01 six nights running. UniFi's scheduled channel optimization moved the radio and the Envoy never rejoined. I turned channel optimization off, and in the six nights after, it didn't happen once.
-
A solar gateway firmware update booked 11 MWh of phantom solar
An Envoy firmware update froze production and rebased the lifetime register, and Home Assistant booked about 11,562 kWh of fake solar in one hour. The CT counter never moved, so I reloaded the integration and rebuilt the statistics from it. An alert now watches the two registers.
-
Nexterm logged my reverse proxy instead of the client
Nexterm logged my reverse proxy as the source of every session because Express wasn't trusting the proxy. I patched it with a bind mounted index.js. Current releases read TRUST_PROXY, so drop the patch and set that to the proxy's own address, not true.
-
I'm fairly sure UniFi threat prevention was eating my traffic
A Proxmox NFS storage check kept timing out on port 111 while 2049 answered, which fits my own UniFi Threat Prevention dropping rpcbind. I didn't add an exception. I switched to a directory storage on the existing mount with is_mountpoint 1, and the timeouts stopped.
-
Upgrading a self-hosted NetBird server without dropping peers
Upgrading my selfhosted NetBird server is pull, recreate, check. The real risk was having no backup, so a weekly timer takes an online SQLite backup, verifies it and alerts on failure. After an upgrade I check for a 200 and a 401, remove orphaned proxy peers, and VACUUM monthly.