Clustering - HA uptime monitor #14

New issue

Closed

opened 2026-02-28 01:31:45 -05:00 by deekerman · 2 comments

deekerman commented

2026-02-28 01:31:45 -05:00

Owner

Originally created by @IIPoliII on GitHub (Jul 12, 2021).

I just saw about this awesome project and really congrats.
It was missing from the opensource side.

The only thing missing by my side is a way to make a cluster with it. Because if it would be possible to run it on 2 servers + at the same time it will assure the server running uptime-kuma will never go down! 😃

Originally created by @IIPoliII on GitHub (Jul 12, 2021). I just saw about this awesome project and really congrats. It was missing from the opensource side. The only thing missing by my side is a way to make a cluster with it. Because if it would be possible to run it on 2 servers + at the same time it will assure the server running uptime-kuma will never go down! 😃

deekerman closed this issue

2026-02-28 01:31:48 -05:00

deekerman commented

2026-02-28 01:31:51 -05:00

Author

Owner

@Stetsed commented on GitHub (Jul 12, 2021):

This isn't really something imo you would integrate into the project, but would use your reverse proxy for(Load balancing), with a check to see which one is up.

If you mean to be able to sync settings etc between the 2 then I might agree this would be useful.

@Stetsed commented on GitHub (Jul 12, 2021): This isn't really something imo you would integrate into the project, but would use your reverse proxy for(Load balancing), with a check to see which one is up. If you mean to be able to sync settings etc between the 2 then I might agree this would be useful.

deekerman commented

2026-02-28 01:31:51 -05:00

Author

Owner

@louislam commented on GitHub (Jul 12, 2021):

Uptime Kuma is a kind of socket server and it is not a stateless service. For my current implementation, It is hard to achieve real clustering.

Docker Swarm + GlusterFS/NFS maybe a good choice if you accept small downtime. Add a uptime-kuma service with 1 replica, and sync the volume with GlusterFS/NFS.

In this setup, when uptime-kuma is down, Docker swarm will quickly spin up another node.

@louislam commented on GitHub (Jul 12, 2021): Uptime Kuma is a kind of socket server and it is not a stateless service. For my current implementation, It is hard to achieve real clustering. Docker Swarm + GlusterFS/NFS maybe a good choice if you accept small downtime. Add a uptime-kuma service with 1 replica, and sync the volume with GlusterFS/NFS. In this setup, when uptime-kuma is down, Docker swarm will quickly spin up another node.

deekerman referenced this issue

2026-02-28 01:50:34 -05:00

Lost every config during update #558

deekerman referenced this issue

2026-02-28 03:28:20 -05:00

Remove monitor warning in logs under maintenance #3406