A hub server is cut off
Weik.io 2026.2One server of a hub with several servers has been out of contact with the others for a minute and refuses changes until it's back.
| Kind | Severity | Who sees it |
|---|---|---|
hub- |
Error | Platform administrators, in a list that isn’t narrowed to an environment or a project |
What it means#
One server of a hub with several servers has lost contact with the others. The changes and the events from agents that reach that server are refused until it’s back in contact.
What raises it#
The leader of the hub’s storage hasn’t heard from a data server for a minute, or a server’s own message server has been out of contact with most of the data servers for a minute. Each backend node checks this itself, without the metric store.
What to check#
See every part of the hub:
$ weikio platform healthOn each hub server, in the stack’s folder, run the checks:
$ docker compose run --rm --no-deps infra weikio-backend doctorHow to fix it#
A server cut off tells a server that is down from a network cut between the servers, and gives the step for each. When the cut ends, the server catches up by itself and the problem clears.