I’m self-hosting Ghost using the Digital Ocean droplet. Working pretty good, but I experience a problem from time to time. Suddently, I get a notification from the uptime monitoring system that my blog is unreachable. When I go to Digital Ocean, I see a pattern like this:
After a ssh to the machine and a
ghost restart everything works as expected.
What could be the problem there?
If you’re looking for some help, it’s important to provide as much context as possible so that people are able to assist you:
- What’s your URL?
- What version of Ghost are you using?
- What configuration?
- What browser?
- What errors or information do you see in the console?
- What steps could someone else take to reproduce the issue you’re having?
Sure, makes sense.
Blog URL: https://blog.elmah.io
Ghost version: 3.13.1
Configuration: don’t know exactly what you mean there. Running in production mode but no custom config besides the hostname and SSL. Cloudflare is running in front.
Browser: Fails on all browsers. It’s the endpoint that doesn’t answer and not related to the browser. Last time it happened, I saw the cloudflare error page saying that the connection from the browser to cloudflare was successful but the connection from cloudflare to Ghost (Digital Ocean) failed.
Console errors: Which console is that? Is that Ghost log files?
Steps to reproduce: That’s the point. I don’t do anything on the Ghost server. It just happens from time to time and a restart resolves it.
Any help would be appreciated here
I get the same issue when doing Ghost updates. It seems that the resource requirements for the update process have grown quite a lot as it kills my entire vps due to cpu usage… I had to upgrade the vps to be able to do updates, even though for normal activity the server works just fine, with very little load. Basically now I have to stop the server, upgrade its resources, run the updates, then tune it back down for normal operations…
The cpu load looks like this:
Basically 0 load when running (serving) the several Ghost site and then 100% cpu (server load 20-30+) and server crash when attempting an update. This is on a 2vcpu 2GB ram vps (t3a.small on AWS).
Good to know. My CPU spike isn’t related to any updated, though. As far as I know, Ghost doesn’t update anything automatically, but this could of course be another internal scheduled task or something?
Unsure what could cause it in your case (could be unrelated to ghost, maybe to nginx or some other component at server level), but my cpu spike is ghost update related.