Home » Mailing lists » Users » NFS with RHEL5 kernel?
NFS with RHEL5 kernel? [message #11913] |
Wed, 11 April 2007 18:41 |
efball
Messages: 41 Registered: September 2006 Location: Santa Rosa, California
|
Member |
|
|
I have NFS mounts inside the VE working fine with
kernel-smp-2.6.18-ovz028stab027.1.i686.rpm.
I tried the ovzkernel-2.6.18-8.el5.028stab027.1.i686.rpm
kernel, but NFS doesn't seem to work.
Is NFS supported in the openvz RHEL5 kernels? The changlog mentions NFS
a couple places, but maybe that's not NFS mounts within the VE?
The NFS page in the wiki seems out of date with the new kernels.
Thank you,
--
E Frank Ball efball@efball.com
|
|
|
|
|
|
Re: NFS with RHEL5 kernel? [message #12910 is a reply to message #11935] |
Mon, 14 May 2007 23:32 |
efball
Messages: 41 Registered: September 2006 Location: Santa Rosa, California
|
Member |
|
|
I tried using the RHEL 5 kernel again
(ovzkernel-2.6.18-8.el5.028stab031.1.i686.rpm) and I still
can't get NFS to work. I tried it without amd (the
automounter) this time. I tried mounts from both HP-UX and
NAS servers.
I get a "mount: permission denied" error, and this spews
onto the console:
lockd_up: no pid, 10 users??
kernel_thread call inside VE
[<c0401b07>] kernel_thread+0x37/0xca
[<c0459d88>] __alloc_pages+0x2bd/0x2d8
[<e0c2da9c>] svc_create_thread+0xe3/0x10b [sunrpc]
[<e0bd2e2a>] lockd+0x0/0x2e2 [lockd]
[<e0bd2bcd>] lockd_up+0x100/0x135 [lockd]
[<e0cf6742>] nfs_start_lockd+0x16/0x33 [nfs]
[<e0cf7296>] nfs_create_server+0x1f8/0x3fe [nfs]
[<c0425816>] __do_softirq+0x84/0x109
[<c04259ec>] local_bh_enable+0x74/0x7f
[<c059f749>] dev_queue_xmit+0x22d/0x24c
[<c05be39d>] ip_output+0x1ec/0x217
[<c0486b25>] __d_lookup+0x98/0xfe
[<c047d4c0>] do_lookup+0x6a/0x178
[<e091892d>] ext3_permission+0x0/0xa [ext3]
[<e0cfd79a>] nfs_get_sb+0x3d5/0x3f5 [nfs]
[<c0477936>] vfs_kern_mount+0x39/0x68
[<c0477997>] do_kern_mount+0x25/0x36
[<c048b8b9>] do_mount+0x618/0x699
[<c04185de>] find_busiest_group+0xeb/0x45c
[<c04386d7>] change_slab_charged+0xce/0xec
[<c041b14b>] rebalance_tick+0x2d2/0x4a9
[<c04367a6>] hrtimer_run_queues+0xd3/0x15e
[<c048a4a7>] copy_mount_options+0x90/0x109
[<e091892d>] ext3_permission+0x0/0xa [ext3]
[<e0cfd79a>] nfs_get_sb+0x3d5/0x3f5 [nfs]
[<c0477936>] vfs_kern_mount+0x39/0x68
[<c0477997>] do_kern_mount+0x25/0x36
[<c048b8b9>] do_mount+0x618/0x699
[<c04386d7>] change_slab_charged+0xce/0xec
[<c0438860>] ub_slab_charge+0x67/0x10a
[<c05c0152>] inet_bind_bucket_create+0x10/0x4b
[<c05c1bf9>] inet_csk_get_port+0x21c/0x226
[<c05f8787>] _spin_lock_bh+0x8/0x18
[<c0596088>] release_sock+0xc/0x91
[<c05dc682>] inet_bind+0x1d6/0x1de
[<c048a4a7>] copy_mount_options+0x90/0x109
[<c0432473>] search_exception_tables+0x14/0x25
[<c0415dab>] fixup_exception+0xb/0x20
[<c05f98c5>] do_page_fault+0x30f/0x498
[<c05f95b6>] do_page_fault+0x0/0x498
[<c05fba29>] iret_exc+0x4e9/0xac0
[<c048b9a7>] sys_mount+0x6d/0xa5
[<c05fa67b>] syscall_call+0x7/0xb
[<c05f007b>] xfrm_get_sa+0x4b/0x6a
=======================
lockd_up: create thread failed, error=-1
This is what I have in /etc/fstab:
lapis:/vol/v1/d27/users/sm2 /home/sm2 nfs
hard,intr,nosuid,actimeo=1,rsize=8192,wsize=8192 0 0
lapis:/vol/v1/d27/SM2 /gfs/prod/SM2 nfs
hard,intr,nosuid,actimeo=1,rsize=8192,wsize=8192 0 0
socobld:/hfs/d1/gos /gfs/prod/gos nfs
hard,intr,nosuid,actimeo=1,rsize=8192,wsize=8192 0 0
beryl:/vol/v8/d22/data/MIDRF1 /gfs/prod/MIDRF1 nfs
hard,intr,nosuid,actimeo=1,rsize=8192,wsize=8192 0 0
E Frank Ball efball@efball.com
On Thu, Apr 12, 2007 at 12:44:45PM +0400, Vasily Tarasov wrote:
> Hello,
>
> Can you describe in more details how it doesn't work? :)
>
> Thanks,
> Vasily.
>
> E Frank Ball wrote:
> >I have NFS mounts inside the VE working fine with
> >kernel-smp-2.6.18-ovz028stab027.1.i686.rpm.
> >
> >I tried the ovzkernel-2.6.18-8.el5.028stab027.1.i686.rpm
> >kernel, but NFS doesn't seem to work.
> >
> >Is NFS supported in the openvz RHEL5 kernels? The changlog mentions NFS
> >a couple places, but maybe that's not NFS mounts within the VE?
> >
> >The NFS page in the wiki seems out of date with the new kernels.
> >
> >Thank you,
> >
>
--
E Frank Ball efball@efball.com
|
|
|
Re: NFS with RHEL5 kernel? [message #12913 is a reply to message #12910] |
Tue, 15 May 2007 06:21 |
Vasily Tarasov
Messages: 1345 Registered: January 2006
|
Senior Member |
|
|
Hello,
you should allow kernel threads in VEs by doing:
HN# sysctl kernel.ve_allow_kthreads=1
Thanks,
Vasily
On Mon, 2007-05-14 at 16:32 -0700, E Frank Ball wrote:
> I tried using the RHEL 5 kernel again
> (ovzkernel-2.6.18-8.el5.028stab031.1.i686.rpm) and I still
> can't get NFS to work. I tried it without amd (the
> automounter) this time. I tried mounts from both HP-UX and
> NAS servers.
>
> I get a "mount: permission denied" error, and this spews
> onto the console:
>
>
> lockd_up: no pid, 10 users??
> kernel_thread call inside VE
> [<c0401b07>] kernel_thread+0x37/0xca
> [<c0459d88>] __alloc_pages+0x2bd/0x2d8
> [<e0c2da9c>] svc_create_thread+0xe3/0x10b [sunrpc]
> [<e0bd2e2a>] lockd+0x0/0x2e2 [lockd]
> [<e0bd2bcd>] lockd_up+0x100/0x135 [lockd]
> [<e0cf6742>] nfs_start_lockd+0x16/0x33 [nfs]
> [<e0cf7296>] nfs_create_server+0x1f8/0x3fe [nfs]
> [<c0425816>] __do_softirq+0x84/0x109
> [<c04259ec>] local_bh_enable+0x74/0x7f
> [<c059f749>] dev_queue_xmit+0x22d/0x24c
> [<c05be39d>] ip_output+0x1ec/0x217
> [<c0486b25>] __d_lookup+0x98/0xfe
> [<c047d4c0>] do_lookup+0x6a/0x178
> [<e091892d>] ext3_permission+0x0/0xa [ext3]
> [<e0cfd79a>] nfs_get_sb+0x3d5/0x3f5 [nfs]
> [<c0477936>] vfs_kern_mount+0x39/0x68
> [<c0477997>] do_kern_mount+0x25/0x36
> [<c048b8b9>] do_mount+0x618/0x699
> [<c04185de>] find_busiest_group+0xeb/0x45c
> [<c04386d7>] change_slab_charged+0xce/0xec
> [<c041b14b>] rebalance_tick+0x2d2/0x4a9
> [<c04367a6>] hrtimer_run_queues+0xd3/0x15e
> [<c048a4a7>] copy_mount_options+0x90/0x109
> [<e091892d>] ext3_permission+0x0/0xa [ext3]
> [<e0cfd79a>] nfs_get_sb+0x3d5/0x3f5 [nfs]
> [<c0477936>] vfs_kern_mount+0x39/0x68
> [<c0477997>] do_kern_mount+0x25/0x36
> [<c048b8b9>] do_mount+0x618/0x699
> [<c04386d7>] change_slab_charged+0xce/0xec
> [<c0438860>] ub_slab_charge+0x67/0x10a
> [<c05c0152>] inet_bind_bucket_create+0x10/0x4b
> [<c05c1bf9>] inet_csk_get_port+0x21c/0x226
> [<c05f8787>] _spin_lock_bh+0x8/0x18
> [<c0596088>] release_sock+0xc/0x91
> [<c05dc682>] inet_bind+0x1d6/0x1de
> [<c048a4a7>] copy_mount_options+0x90/0x109
> [<c0432473>] search_exception_tables+0x14/0x25
> [<c0415dab>] fixup_exception+0xb/0x20
> [<c05f98c5>] do_page_fault+0x30f/0x498
> [<c05f95b6>] do_page_fault+0x0/0x498
> [<c05fba29>] iret_exc+0x4e9/0xac0
> [<c048b9a7>] sys_mount+0x6d/0xa5
> [<c05fa67b>] syscall_call+0x7/0xb
> [<c05f007b>] xfrm_get_sa+0x4b/0x6a
> =======================
> lockd_up: create thread failed, error=-1
>
>
> This is what I have in /etc/fstab:
>
> lapis:/vol/v1/d27/users/sm2 /home/sm2 nfs
> hard,intr,nosuid,actimeo=1,rsize=8192,wsize=8192 0 0
>
> lapis:/vol/v1/d27/SM2 /gfs/prod/SM2 nfs
> hard,intr,nosuid,actimeo=1,rsize=8192,wsize=8192 0 0
>
> socobld:/hfs/d1/gos /gfs/prod/gos nfs
> hard,intr,nosuid,actimeo=1,rsize=8192,wsize=8192 0 0
>
> beryl:/vol/v8/d22/data/MIDRF1 /gfs/prod/MIDRF1 nfs
> hard,intr,nosuid,actimeo=1,rsize=8192,wsize=8192 0 0
>
>
> E Frank Ball efball@efball.com
>
>
>
> On Thu, Apr 12, 2007 at 12:44:45PM +0400, Vasily Tarasov wrote:
> > Hello,
> >
> > Can you describe in more details how it doesn't work? :)
> >
> > Thanks,
> > Vasily.
> >
> > E Frank Ball wrote:
> > >I have NFS mounts inside the VE working fine with
> > >kernel-smp-2.6.18-ovz028stab027.1.i686.rpm.
> > >
> > >I tried the ovzkernel-2.6.18-8.el5.028stab027.1.i686.rpm
> > >kernel, but NFS doesn't seem to work.
> > >
> > >Is NFS supported in the openvz RHEL5 kernels? The changlog mentions NFS
> > >a couple places, but maybe that's not NFS mounts within the VE?
> > >
> > >The NFS page in the wiki seems out of date with the new kernels.
> > >
> > >Thank you,
> > >
> >
>
|
|
|
Re: NFS with RHEL5 kernel? [message #12926 is a reply to message #12913] |
Tue, 15 May 2007 08:09 |
ldv
Messages: 24 Registered: June 2006
|
Junior Member |
|
|
Hi,
On Tue, May 15, 2007 at 10:21:56AM +0400, Vasily Tarasov wrote:
> you should allow kernel threads in VEs by doing:
>
> HN# sysctl kernel.ve_allow_kthreads=1
Anyway, current kernel diagnostics on this subject is too verbose, isn't it?
--
ldv
|
|
|
Goto Forum:
Current Time: Thu Oct 17 08:41:06 GMT 2024
Total time taken to generate the page: 0.05388 seconds
|