Suggestion: Status page #15

jdoubleu commented 2019年01月01日 20:35:14 +01:00

You may provide a status page for all service using FOSS like http://cachethq.io/?

You may provide a status page for all service using FOSS like http://cachethq.io/?

hw commented 2019年01月01日 20:43:16 +01:00

This has been requested before, we are looking into this. Current status monitors are based on netdata, and potentially leaking details that could be exploited for DDOS etc.; a more coarse-grained view could indeed be useful.

This has been requested before, we are looking into this. Current status monitors are based on netdata, and potentially leaking details that could be exploited for DDOS etc.; a more coarse-grained view could indeed be useful.

hw commented 2019年01月02日 20:38:20 +01:00

FYI https://alt-three.com/apilayer-have-acquired-cachet/

FYI https://alt-three.com/apilayer-have-acquired-cachet/

ntimo commented 2019年01月21日 22:25:53 +01:00

What about a public Grafana dashboard like the one from GitLab? https://dashboards.gitlab.com/

What about a public Grafana dashboard like the one from GitLab? https://dashboards.gitlab.com/

davidak commented 2019年01月22日 23:26:40 +01:00

@ntimo too verbose for users that just want to know if the site is DOWN.

We have netdata for monitoring.

@ntimo too verbose for users that just want to know if the site is DOWN. We have [netdata](https://github.com/netdata/netdata) for monitoring.

hjacobs commented 2019年03月23日 12:03:23 +01:00

Related: #36

Related: #36

Ghost commented 2019年04月08日 11:40:14 +02:00

We use https://staytus.co/, as an Alternative to Cachet, and removed Third Party Requests.
Maybe it's worth consideration?

We use https://staytus.co/, as an Alternative to Cachet, and removed Third Party Requests. Maybe it's worth consideration?

jdoubleu commented 2019年04月09日 09:52:54 +02:00

I also found cState (https://github.com/cstate/cstate). Example: https://status.uberspace.de/

I also found cState (https://github.com/cstate/cstate). Example: https://status.uberspace.de/

hw commented 2019年04月11日 08:16:43 +02:00

cState looks nice, and fits very well into the current workflow: updating posts as markdown in git is pretty much exactly the same as we do the blog...

cState looks nice, and fits very well into the current workflow: updating posts as markdown in git is pretty much exactly the same as we do the blog...

6543 commented 2020年06月25日 01:37:18 +02:00

any news?

any news?

hw commented 2020年06月25日 09:17:02 +02:00

Anybody willing to contribute a deployment setup? (potentially a script installing+setting up all packages needed on a minimum-cost instance and easy to migrate if needed. Basically something that takes IP and username, and configures everything needed via ssh only).

Anybody willing to contribute a deployment setup? (potentially a script installing+setting up all packages needed on a minimum-cost instance and easy to migrate if needed. Basically something that takes IP and username, and configures everything needed via ssh only).

fnetX commented 2022年01月19日 02:48:35 +01:00

We are currently running a public status page at https://status.codeberg-test.org. If this works out, we'll migrate it to the main domain.

We are currently running a public status page at https://status.codeberg-test.org. If this works out, we'll migrate it to the main domain.

davidak commented 2022年02月22日 21:30:24 +01:00

@fnetX did you encounter any issues or can we proceed to use it?

@fnetX did you encounter any issues or can we proceed to use it?

fnetX commented 2022年02月22日 21:32:54 +01:00

Well, it's acting weird to be honest. It's having more hickup than Codeberg does, and most if not all current notices are false positives.

Also, the stats are not very useful as it does only consider some very recent data. Uptime is usually at 100%, decreasing significantly after an issue and going back up to 100% very soon.

@rwa maybe you can give me admin access so I can play around with the dashboard for myself. Other than that, the most important missing piece is the DNS and haproxy setup, I think this is best done by @hw.

Well, it's acting weird to be honest. It's having more hickup than Codeberg does, and most if not all current notices are false positives. Also, the stats are not very useful as it does only consider some very recent data. Uptime is usually at 100%, decreasing significantly after an issue and going back up to 100% very soon. @rwa maybe you can give me admin access so I can play around with the dashboard for myself. Other than that, the most important missing piece is the DNS and haproxy setup, I think this is best done by @hw.

davidak commented 2022年02月22日 21:38:13 +01:00

Can you report the issues upstream? I would not give up on the project yet.

Can you report the issues upstream? I would not give up on the project yet.

fnetX commented 2022年02月22日 21:47:01 +01:00

No, me neither, but I'm also not considering it ready for public launch if we might need to replace it.

Some things are also somewhat nice, e.g. the fact that the TLS cert for codeberg.page is expiring within seven days and not renewed yet (CC @momar will the pages server take care of this? 🧐)

No, me neither, but I'm also not considering it ready for public launch if we might need to replace it. Some things are also somewhat nice, e.g. the fact that the TLS cert for codeberg.page is expiring within seven days and not renewed yet (CC @momar will the pages server take care of this? 🧐)

Ghost commented 2022年02月22日 21:52:02 +01:00

Also, the stats are not very useful as it does only consider some very recent data. Uptime is usually at 100%, decreasing significantly after an issue and going back up to 100% very soon.

This is due to the fact that the uptime on the status page is calculated for 24 hours only, which seems a bit arkward. Something like 30 days or so seems more reasonable for us.

Afaik this is not customizable by a setting currently, may be worth an issue/PR upstream.

and most if not all current notices are false positives.

That seems weird, i have a ptrivate uptime kuma instance running and it works pretty decent. Maybe i can have a look in the next days whats going on.

I'll send you the credentials via Matrix.

> Also, the stats are not very useful as it does only consider some very recent data. Uptime is usually at 100%, decreasing significantly after an issue and going back up to 100% very soon. This is due to the fact that the uptime on the status page is calculated for 24 hours only, which seems a bit arkward. Something like 30 days or so seems more reasonable for us. Afaik this is not customizable by a setting currently, may be worth an issue/PR upstream. > and most if not all current notices are false positives. That seems weird, i have a ptrivate uptime kuma instance running and it works pretty decent. Maybe i can have a look in the next days whats going on. I'll send you the credentials via Matrix.

Ghost commented 2022年10月25日 12:05:38 +02:00

The status page (powered by "Uptime Kuma") is now officially available at https://status.codeberg.org/

The status page (powered by "Uptime Kuma") is now officially available at https://status.codeberg.org/