All of lore.kernel.org
 help / color / mirror / Atom feed
* strange behavior with nfsv3 using linux v4.1
@ 2015-11-06 22:57 William Dauchy
  2015-11-07  8:41 ` William Dauchy
  0 siblings, 1 reply; 3+ messages in thread
From: William Dauchy @ 2015-11-06 22:57 UTC (permalink / raw)
  To: Linux NFS mailing list; +Cc: Trond Myklebust

[-- Attachment #1: Type: text/plain, Size: 2993 bytes --]

Hello,

I am seeing strange behaviors with a single nfsv3 mount on linux
v4.1.x. I first started to see two kworker in D state while I was not
doing anything. The server is an opensolaris nfs server.

If I try to get more info I am seeing getattr loop as follow:

echo 65535 > /proc/sys/sunrpc/nfs_debug

NFS call  getattr
NFS reply getattr: 0
NFS: nfs_update_inode(0:15/22630 fh_crc=0x1df1ef00 ct=1 info=0x27e7f)
NFS: (0:15/22630) revalidation complete
NFS: nfs_lookup_revalidate(x86_64-linux-gnu/libc-2.19.so) is valid
NFS: permission(0:15/22630), mask=0x24, res=0
NFS: open file(x86_64-linux-gnu/libc-2.19.so)
NFS: dentry_delete(x86_64-linux-gnu/libc.so.6, 6808cc)
NFS: read(x86_64-linux-gnu/libc-2.19.so, 832@0)
NFS: mmap(x86_64-linux-gnu/libc-2.19.so)
NFS: mmap(x86_64-linux-gnu/libc-2.19.so)
NFS: flush(x86_64-linux-gnu/libc-2.19.so)
NFS: permission(0:15/4), mask=0x81, res=0
NFS: nfs_lookup_revalidate(/etc) is valid
NFS: permission(0:15/12), mask=0x81, res=0
NFS: nfs_lookup_revalidate(etc/ld.so.nohwcap) is valid
NFS: permission(0:15/4), mask=0x81, res=0
NFS: nfs_lookup_revalidate(/lib) is valid
NFS: permission(0:15/14), mask=0x81, res=0
NFS: nfs_lookup_revalidate(lib/x86_64-linux-gnu) is valid
NFS: permission(0:15/1178), mask=0x81, res=0
NFS: nfs_lookup_revalidate(x86_64-linux-gnu/libpthread.so.0) is valid
NFS: permission(0:15/1178), mask=0x1, res=0
NFS: revalidating (0:15/22644)
NFS call  getattr
NFS reply getattr: 0
NFS: nfs_update_inode(0:15/22644 fh_crc=0x9a566cc1 ct=1 info=0x27e7f)
NFS: (0:15/22644) revalidation complete
NFS: nfs_lookup_revalidate(x86_64-linux-gnu/libpthread-2.19.so) is valid
NFS: permission(0:15/22644), mask=0x24, res=0
NFS: open file(x86_64-linux-gnu/libpthread-2.19.so)
NFS: dentry_delete(x86_64-linux-gnu/libpthread.so.0, 6808cc)
NFS: read(x86_64-linux-gnu/libpthread-2.19.so, 832@0)
NFS: mmap(x86_64-linux-gnu/libpthread-2.19.so)
NFS: mmap(x86_64-linux-gnu/libpthread-2.19.so)
NFS: flush(x86_64-linux-gnu/libpthread-2.19.so)
NFS: release(etc/ld.so.cache)
NFS: dentry_delete(etc/ld.so.cache, 4808cc)
NFS: permission(0:15/4), mask=0x81, res=0
NFS: nfs_lookup_revalidate(/root) is valid
NFS: permission(0:15/20), mask=0x81, res=0
NFS: nfs_lookup_revalidate(root/.config) is valid
NFS: permission(0:15/4), mask=0x81, res=0
NFS: nfs_lookup_revalidate(/etc) is valid
NFS: permission(0:15/12), mask=0x81, res=0
NFS: nfs_lookup_revalidate(etc/terminal-colors.d) is valid
NFS: permission(0:15/4), mask=0x81, res=0
NFS: nfs_lookup_revalidate(/dev) is valid
NFS: permission(0:15/4), mask=0x81, res=0
NFS: nfs_lookup_revalidate(/var) is valid
NFS: permission(0:15/27), mask=0x81, res=0
NFS: nfs_lookup_revalidate(var/lib) is valid
NFS: permission(0:15/19882), mask=0x81, res=0
NFS: revalidating (0:15/19990)
[...]


Do you have some hints in order to help debug this issue?
If I reboot using my older v3.14.x with same config, I don't have that
behavior.

Best regards,
-- 
William

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: strange behavior with nfsv3 using linux v4.1
  2015-11-06 22:57 strange behavior with nfsv3 using linux v4.1 William Dauchy
@ 2015-11-07  8:41 ` William Dauchy
  2015-11-11 21:28   ` William Dauchy
  0 siblings, 1 reply; 3+ messages in thread
From: William Dauchy @ 2015-11-07  8:41 UTC (permalink / raw)
  To: William Dauchy; +Cc: Linux NFS mailing list, Trond Myklebust

[-- Attachment #1: Type: text/plain, Size: 400 bytes --]

On Nov06 23:57, William Dauchy wrote:
> I am seeing strange behaviors with a single nfsv3 mount on linux
> v4.1.x. I first started to see two kworker in D state while I was not
> doing anything. The server is an opensolaris nfs server.

forgot to mention this nfsv3 mount is a using CONFIG_ROOT_NFS
it's a read only mount point

the client seems to be looping on revalidation.
-- 
William

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: strange behavior with nfsv3 using linux v4.1
  2015-11-07  8:41 ` William Dauchy
@ 2015-11-11 21:28   ` William Dauchy
  0 siblings, 0 replies; 3+ messages in thread
From: William Dauchy @ 2015-11-11 21:28 UTC (permalink / raw)
  To: William Dauchy; +Cc: Linux NFS mailing list, Trond Myklebust

[-- Attachment #1: Type: text/plain, Size: 524 bytes --]

On Nov07 09:41, William Dauchy wrote:
> On Nov06 23:57, William Dauchy wrote:
> > I am seeing strange behaviors with a single nfsv3 mount on linux
> > v4.1.x. I first started to see two kworker in D state while I was not
> > doing anything. The server is an opensolaris nfs server.
> 
> forgot to mention this nfsv3 mount is a using CONFIG_ROOT_NFS
> it's a read only mount point
> 
> the client seems to be looping on revalidation.

please forget this report I mixed it with a network card issue.
-- 
William

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2015-11-11 21:28 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-11-06 22:57 strange behavior with nfsv3 using linux v4.1 William Dauchy
2015-11-07  8:41 ` William Dauchy
2015-11-11 21:28   ` William Dauchy

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.