cgroup support

[alpha_one_x86]

Hello,

I'm very interested by zabbix, but for lxc support I need support of cgroup/namesapce. All graph in one to see quickly the vm in problem (here you have 4 vm: catchchallenger-db-base, catchchallenger-db-common-1-a-master, catchchallenger-db-common-2, catchchallenger-db-login):




Why do you need this? To isolate quickly the group of process responsive of the load, include the virtual machine.
How? Detect the cgroup mount point and use it.

Cheers,

[jan.garaj]

https://github.com/monitoringartist/zabbix-docker-monitoring

- ignore "docker" term and you have monitoring base on cgroup pseudofiles - at least:

docker.mem[cid,mmetric]
docker.cpu[cid,cmetric]
docker.dev[cid,bfile,bmetric]

Are you sure about cgroup in OpenVZ?

https://support.zabbix.com/browse/ZBXNEXT-2460

[alpha_one_x86]

Are you sure about cgroup in OpenVZ?

No, unsure to OpenVZ

[alpha_one_x86]

These don't match with my need, see above, one graphic by item to see quickly what vm use the ressources.

[jan.garaj]

Wrong question => wrong answer. Be specific and also try to use correct terms. Your topic is cgroup, but you have mentioned OpenVZ. Also VM doesn't use cgroups. I would like to see which monitoring tool is able to measure CPU load per cgroup resource :-) You are mixing a lot of technologies and you are asking about monitoring of this mixed universe. Be specific otherwise response will be wrong.

[alpha_one_x86]

Updated graphic to be more clear. With this graphic I see quickly what vm is the problem.

[alpha_one_x86]

UP, any news for this?

[alpha_one_x86]

https://github.com/monitoringartist/zabbix-docker-monitoring

Here you have my work, what guide use to adapt to zabbix?

https://github.com/alphaonex86/contrib/tree/master/plugins/lxc

[jan.garaj]

What is stopping you to use https://github.com/monitoringartist/...ker-monitoring ?