December 24, 20187 yr Happy Holidays folks Recently I made a few changes to my rig and now CPU 0 or 6 (physical/hyper) which are reserved for Unraid goes to 100% after some period of time and becomes locked at that percent. The changes: 1 - Swapped a ASM1142/ASM1543 pci card for a ASM1142 card as the 1543 was causing reboot rather than powerdown conditions - This is stubbed out just like the old card (along with a Sata – both for Windows passthrough) via vfio-pci.ids=1b21:0612,1b21:1242 2 – Upgraded 250 GB SSD to 1T unit as per https://wiki.unraid.net/Replace_A_Cache_Drive 3 – Removed isolcpus=3-5,9-11 stub as it is no longer required with GUI built in pinning TOP command shows that *kworker/6:2+pm* is consuming 83% of the available CPU Logs attached. Thanks ahead of time for looking! tower-diagnostics-20181224-0827.zip Edited December 24, 20187 yr by landS
December 29, 20187 yr Author Rebooted server after first post. Happened again within the last 24 hours - 1 thread is stuck at 100% (kworker/6:3+pm). Fresh logs attached. Any help would be appreciated. tower-diagnostics-20181229-0602.zip Edited December 29, 20187 yr by landS
January 3, 20197 yr Author Hopefully I'm not speaking too soon, but originally passed through a Logitech unified usb receiver to a VM. This usb device has failed and no longer shows up on the list - when attempting the VM start it gave a USB failure message and device id. Inserted a new Logitech unified receiver and it has the same device ID as the original. Since doing this, no more pegged CPU.
January 5, 20197 yr Author Spoke too soon. 1 thread stuck again at 100%. Any ideas from the prior logs folks? Thanks for taking a gander.
February 6, 20206 yr google "linux kworker 100% cpu" there are many forum posts about this with various things to try.
Archived
This topic is now archived and is closed to further replies.