Hello folks !
I looked over the forum and did not found any solution so I’m asking here.
I have a bare metal server with 126Gb of Memory, properly detected by Netdata, cloud instance of Netdata, and I’m receiving alert of type “90% cgroup_ram_in_use” when used Memory is around 15Gb , whereas I have way more available memory.
How to fix/increase the memory limit detected by NetData ?
I check the file /etc/netdata/edit-config health.d/cgroups.conf and saw this line
calc: ($ram) * 100 / $memory_limit
But do not know where this “memory_limit” comes from .
Thanks a lot for your help !
This alert isn’t about your bare metal server, but about a VM or container running on it (these are the typical examples of Unix cgroups). So I expect Netdata is telling you than one of your VMs is running out of memory. If that’s the case (you have many VMs running on that server), I suggest you:
- Install Netdata on each VM as well
- Set up replication to have the Netdata agents running on their VMs stream their metrics to the Netdata running on the bare metal server
Hi @Christopher_Akritid1 ,
thanks a lot for taking the time to answer.
This make totally sense, I’m hosting several docker instances on it.
Investigation closed !
Wishing you a nice day