SOLVED: SMBD using 100% of available CPU
Posted: Thu Nov 19, 2015 8:20 am
OK first the basics:
Running Amahi 8, fully updated
smbd is replicating numerous times (in system monitor I have over 100 of them) and all 5 of my CPU are pegged at 100%
If I go into the hda dashboard and stop the samba server (or restart it) cpu usage drops to about 3% (and obviously all the smbd dameons go away)
This happens even with the amahi server is disconnected from the network, so nothing external is spawning those processes.
shares are mounted locally
Greyhole is on
I have 19 hard drives in the greyhole pool
More detailed version:
I'm running Amahi8 on a VM, within the last week or so it has started using 100% of however many CPU (up to 10x2.4 ghz per processor) I've thrown at it for samba. This happens even when I disable eth0 within fedora so all those request are spawning locally.
File system browsing is either REALLY slow within the shares (ls on a single directory taking 10+ minutes) or fails with a "resource unavailable" message. running an ls command on locations not in the shares/greyhole works quickly and as expected.
Last night I ran fsck on all disks in greyhole, with no errors found, so I don't think it's a hardware problem. The OS disks and drives in greyhole share are all in the same location and connected through the same controller cards so I don't believe that a failure there is the issue since then they'd have the same slow downs.
The smbd process seems to spawn about 1 per every 2 minutes
If this is the wrong location, please either move the post, or tell me where I should repost.
Running Amahi 8, fully updated
smbd is replicating numerous times (in system monitor I have over 100 of them) and all 5 of my CPU are pegged at 100%
If I go into the hda dashboard and stop the samba server (or restart it) cpu usage drops to about 3% (and obviously all the smbd dameons go away)
This happens even with the amahi server is disconnected from the network, so nothing external is spawning those processes.
shares are mounted locally
Greyhole is on
I have 19 hard drives in the greyhole pool
More detailed version:
I'm running Amahi8 on a VM, within the last week or so it has started using 100% of however many CPU (up to 10x2.4 ghz per processor) I've thrown at it for samba. This happens even when I disable eth0 within fedora so all those request are spawning locally.
File system browsing is either REALLY slow within the shares (ls on a single directory taking 10+ minutes) or fails with a "resource unavailable" message. running an ls command on locations not in the shares/greyhole works quickly and as expected.
Last night I ran fsck on all disks in greyhole, with no errors found, so I don't think it's a hardware problem. The OS disks and drives in greyhole share are all in the same location and connected through the same controller cards so I don't believe that a failure there is the issue since then they'd have the same slow downs.
The smbd process seems to spawn about 1 per every 2 minutes
If this is the wrong location, please either move the post, or tell me where I should repost.