francescocariati
Cadet
- Joined
- Apr 2, 2018
- Messages
- 6
Hi,
I'm a FreeNAS user since 2011, on various machines.
I'm experiencing an high load average on a freshly installed 11.1-U4
on an HPE MicroServer Gen 10, Opteron X3216 dual core, 8 GB RAM ECC (I know it's low, I'm still waiting for the extra sticks to arrive)
UEFI Boot is from a Kingston SSD, storage is on 2 x 2TB WD SATA, RAID 1.
I was not experiencing any similar issue when running on 9.10.x
Despite the fact that the machine has zero traffic (and therefore zero disk activity) coming from the users, you can see that the load average is quite high and has a lot of spikes:
The culprit seems to be a bunch of python processes
While middlewared is always active, I think the periodic high-CPU spikes are caused by the autosnap.py process which starts (and ends) every minute.
I've done some research on the Issue Tracker and while the issue has been reported recently, it's unclear to me if there's a stable workaround (like a patched autosnap.py) or if the mechanism will be changed in a later release of FreeNAS:
https://redmine.ixsystems.com/issues/25757
https://redmine.ixsystems.com/issues/27478
Can you please help me?
Thanks in advance
I'm a FreeNAS user since 2011, on various machines.
I'm experiencing an high load average on a freshly installed 11.1-U4
on an HPE MicroServer Gen 10, Opteron X3216 dual core, 8 GB RAM ECC (I know it's low, I'm still waiting for the extra sticks to arrive)
UEFI Boot is from a Kingston SSD, storage is on 2 x 2TB WD SATA, RAID 1.
I was not experiencing any similar issue when running on 9.10.x
Despite the fact that the machine has zero traffic (and therefore zero disk activity) coming from the users, you can see that the load average is quite high and has a lot of spikes:

The culprit seems to be a bunch of python processes

While middlewared is always active, I think the periodic high-CPU spikes are caused by the autosnap.py process which starts (and ends) every minute.
I've done some research on the Issue Tracker and while the issue has been reported recently, it's unclear to me if there's a stable workaround (like a patched autosnap.py) or if the mechanism will be changed in a later release of FreeNAS:
https://redmine.ixsystems.com/issues/25757
https://redmine.ixsystems.com/issues/27478
Can you please help me?
Thanks in advance