Forums

Home / Forums

You need to log in to create posts and topics. Login · Register

Passed queue_depth: 512 exceeds TCM/SE_Device MAX TCQ: 256

I am seeing this message in all nodes of my cluster , only one time as of now.

I have been reading it might be caused by the queue length from controller or disk , but quite lost on this.

Is it something I should worry about ? Any idea how to solve it?

look in the config file:
/opt/petasan/config/tuning/current/lio_tunings
if under "storage_objects": it shows
"queue_depth": 512
change it to
"queue_depth": 256

you should move the paths to other nodes then back, this will create a new connection with updated queue depth

Hi , I tried this , but path reassignment is always failing.


















 

Thanks,

After changing the lio tunings, can you reboot

Sure,  then I should restart all nodes one by one to apply this settings.

About the path re-assignment, any idea why is not working ?

i suspect it is related to the same thing, let me know if the issue persists.

Just applied all tuning settings to all nodes, the error message about queue_depth disappeared. I am opening a new post about path the reassignment issue as this one got solved.

Thx!