Skip to contentSkip to conversation

New: Problem with NFS kernel server code

Bug 208807 Stephen Hemminger 2 participants 2 messages Aug 4, 2020

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

Begin forwarded message:

Date: Tue, 04 Aug 2020 11:33:23 +0000
From: [email protected]
To: [email protected]
Subject: [Bug 208807] New: Problem with NFS kernel server code

https://bugzilla.kernel.org/show_bug.cgi?id=208807

            Bug ID: 208807
           Summary: Problem with NFS kernel server code
           Product: Networking
           Version: 2.5
    Kernel Version: 5.8.0
          Hardware: All
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: Other
          Assignee: [email protected]
          Reporter: [email protected]
        Regression: No

Created attachment 290759
--> https://bugzilla.kernel.org/attachment.cgi?id=290759&action=edit
Output of dmesg while machine was dying

After about an hour and a half of operating on 5.8.0, one of our NFS servers began to slow and basically ground to a halt. Looks like a spinlock issue in the NFS kernel server code. I'll attach the output of dmesg while the machine was dying in a file called "spinlock". A shame because while it was running, the performance of 5.8 was astounding. Only the NFS servers seem to get ill, the
client machines seem to run fine on 5.8 so far at least.

Signature · 2 lines
You are receiving this mail because:
You are the assignee for the bug.
#1 lore raw

Commented, but I guess it should really be reassigned to nfsd, I'm not sure how.

--b.

Quoted 43 lines · Stephen Hemminger

Begin forwarded message:

Date: Tue, 04 Aug 2020 11:33:23 +0000
From: [email protected]
To: [email protected]
Subject: [Bug 208807] New: Problem with NFS kernel server code

https://bugzilla.kernel.org/show_bug.cgi?id=208807

            Bug ID: 208807
           Summary: Problem with NFS kernel server code
           Product: Networking
           Version: 2.5
    Kernel Version: 5.8.0
          Hardware: All
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: Other
          Assignee: [email protected]
          Reporter: [email protected]
        Regression: No

Created attachment 290759
--> https://bugzilla.kernel.org/attachment.cgi?id=290759&action=edit
Output of dmesg while machine was dying

After about an hour and a half of operating on 5.8.0, one of our NFS servers began to slow and basically ground to a halt. Looks like a spinlock issue in the NFS kernel server code. I'll attach the output of dmesg while the machine was dying in a file called "spinlock". A shame because while it was running, the performance of 5.8 was astounding. Only the NFS servers seem to get ill, the
client machines seem to run fine on 5.8 so far at least.

Signature · 2 lines
You are receiving this mail because:
You are the assignee for the bug.