Good day,
How can I obtain information on especially which upstream(s) was dialed and typically the response times?
and the headers?
I thought I saw it in some caddylog, but currently can’t see it.
Good day,
How can I obtain information on especially which upstream(s) was dialed and typically the response times?
and the headers?
I thought I saw it in some caddylog, but currently can’t see it.
You can use the log directive to enable access logs:
To get more details about proxied requests, you can enable the debug
global option which reveals some additional information:
Ah! Thank you, was using the JSON config, and it seems I have to include “http.handlers.reverse_proxy” to show that with the access_logs:
curl -s http://localhost:2019/config/logging/logs/log0/ | jq .
{
"encoder": {
"format": "json"
},
"include": [
"http.handlers.reverse_proxy",
"http.log.access.log0"
],
"level": "DEBUG",
"writer": {
"filename": "/var/log/caddy/caddy.log",
"output": "file",
"roll_keep": 30,
"roll_keep_days": 14,
"roll_size_mb": 10
}
}
Is it possible, to (like the http.log.access. ) to name each upstream/dial/proxy separately?
You mean the log names? No, but since they’re structured you can just ignore the ones you don’t want.
Hmmm… okay, I also see that reverse_proxy only emit logs on DEBUG level.
the reason I’d like to have log names associated with an upstream, is that way I could easier group the various upstreams and their http access logs together in the same files for the clients, instead of now having to first filter for them the reverse_proxy.
This topic was automatically closed after 30 days. New replies are no longer allowed.