Moved install to new server - don't think API is working

[garlandtech]

I just moved my LibreNMS install from a Centos7 to Ubuntu 22 server. My old server had a fully functioning (well I thought it was, but there may have been issues) install of Oxidized running.

I have re-installed Oxidized and put the server back together per the docs and what I feel was common sense. Everything on the LibreNMS side seems to be working, but the Oxidized is tossing a 404 error when starting causing it to crash. It “:seems” like the API for Oxidized isn’t “up” or “listening”. Is there a way to check that?

Would Ubuntu Nginx have a different “something” that I need to account for for the copy?
Things I’ve done:

• Installed latest Oxidized per docs, and copied the original config file over
• copied the original GIT repository which should be sufficient
• Created a new API key for Oxidied and entered into the config file

What I’m seeing:
oxidized@gti-librenms:~$ oxidized
I, [2023-09-11T20:55:04.133540 #987795] INFO – : Oxidized starting, running as pid 987795
I, [2023-09-11T20:55:04.384941 #987795] INFO – : lib/oxidized/nodes.rb: Loading nodes
F, [2023-09-11T20:55:04.388660 #987795] FATAL – : Oxidized crashed, crashfile written in /home/oxidized/.config/oxidized/crash
809: unexpected token at '

Any help would be greatly appreciated!
Thank you,
Sean

[RR1]

you can check librems api for oxidized status with below curl query

curl -H ‘X-Auth-Token: YOURTOKEN’ http://librenms_url/api/v0/oxidized

it should list devices in JSON format

[garlandtech]

Thank you RR1,
I setup LetsEncrypt with the new server (same as old, but I think old was Apache on Centos 7), and it has some rewrite rules for port 80 traffic… I changed the server API path in Oxidized to the FQDN with https and BAM it’s working now.

Thank you very much! I used the curl command to test various iterations until it returned the list of devices