SCALE Overall System Stability Issues
Description
Problem/Justification
Impact
clones
SmartDraw Connector
Katalon Manual Tests (BETA)
Activity

Ryan Moeller June 22, 2022 at 1:55 PM
Feedback timeout.

Ryan Moeller January 21, 2022 at 2:57 PM
The load averages in the htop screenshot look pretty high to me. The confusion is because top/htop do not show kernel threads by default.
The stack trace in dmesg is a side effect of the system being under heavy load.
The debug does not show the system under load, so I cannot tell what may be going on from it. Please provide the output of ps -auxwwf while the system is exhibiting this behavior so we can see which kernel threads are busy.

cyrus104 May 1, 2021 at 2:14 PM
Yes, it did. I exported the datasets from CORE and then used the import wizard in SCALE.

Andrew Walker May 1, 2021 at 12:27 PM
acltype on SMB shares is off. Did these datasets originate from FreeBSD / Core?

cyrus104 May 1, 2021 at 5:27 AMEdited
I cannot figure out how to edit the main text body.
Right now my webgui still "works" but whenever I try to make a change or do anything it freezes and just spins forever. I have tested this on the latest Chrome and Firefox.
I think I need to reboot the machine but I have several XCP-ng VMs remotely running on it plus some CIFS/NFS datasets that are being used by the those VMs so it's hard to just migrate them to local storage temporarily.
With the webgui frozen, how can I dump the debug info from command line?
Here is the latest DMESG that might apply:
Details
Details
Assignee

Reporter

When I got to the dashboard on SCALE, I get all threads at very high cpu utilization but when I look at htop and top, I don't see high cpu utilization at all. The two picture were taken at the same time. The only thing going on is one SMB file transfer... which instead of going the normal 700MB/s it's going 153MB/s so something is wrong but not sure what.
I am trying to dump the debug but that fails half the time or as right now is stuck at 81% Dump ZFS configuration.