I cannot seem to find any useful info on the topic.
Furthermore, what is the best way to monitor CoreOS (we use observium).
I cannot seem to find any useful info on the topic.
Furthermore, what is the best way to monitor CoreOS (we use observium).
If standard Linux SNMP metrics are most you need, you just want to deploy a container that runs SNMP daemon. For that purpose you will probably need to expose it to host's network namespace (--net=host
, if you are using Docker) and then you definitely need to bind-mount /proc
(with -v /proc:/hostproc
passed to docker run
). The only unusual thing you would need point the SNMP daemon at is the /hostproc
instead of /proc
.
You can probably try using the polinux/snmpd
image for this purpose.