XCP-NG: Help with alarms

Hey all,

I just recently checked up on my server and found that I had around 1100+ alarms and they are all about mem_usage and log_fs_usage. Googling netted nothing conclusive. I am wondering if this is anything to worry about. Below are what the alarms tell me;

mem_usage 0.953628 (Issue on: Control domain on host: navxcp-ng)

alarm_trigger_level 0.95

alarm_trigger_period 60

alarm_auto_inhibit_period 3600

AND

log_fs_usage 1.000000 (Issue on: Control domain on host: navxcp-ng)

alarm_trigger_level 0.9

alarm_trigger_period 60

alarm_auto_inhibit_period 3600

I have not seen that either, I would try their forums https://xcp-ng.org/forum/

Honestly you may want to take a look at the memory you have allocated for Dom0. That one usually is where most of my alarms came from other than over memory utilization and storage.

When you leave the Dom0 to be managed automatically, thats where it won’t scale up as necessary sometimes. I’ve got it in my mind now to just set aside 16 GB of ram just for Dom0. Like if I order a new server, that’s just dedicated 16 Gb just for Dom0 and then on top my requirements. Right now if you have 4 Gb you may want to consider bumping it to 8 Gb.

Hi - ’ log_fs_usage’ probably means your log directory is full, as it is probably set up as its own partition. Login via SSH or use the host console in XOA and run the command df -h and review the output. Look for whatever filesystem is mounted on /var/log. Look at the ‘use’ column, it is probably 100% or quite near it. CD into /var/log and remove some of the old and compressed files. rm -f *.gz and also rm -f *.1 will do this. Then run df -h again and check the use figure went down. This is likely a temporary fix, probably more digging is required to find out what filled up the logs in the first place it this keeps reoccurring…