nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped... - NFS

This is a discussion on nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped... - NFS ; Greetings All, Seeing a lot of nsm_mon_unmon: rpc failed, status=-13 kernel: lockd: cannot monitor 192.168.1.202 where 202 is the client, and 200 is the server. For the server: rpcinfo -p k8ne program vers proto port 100000 2 tcp 111 portmapper ...

+ Reply to Thread
Results 1 to 7 of 7

Thread: nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped...

  1. nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped...

    Greetings All,

    Seeing a lot of

    nsm_mon_unmon: rpc failed, status=-13
    kernel: lockd: cannot monitor 192.168.1.202

    where 202 is the client, and 200 is the server.

    For the server:
    rpcinfo -p k8ne
    program vers proto port
    100000 2 tcp 111 portmapper
    100000 2 udp 111 portmapper
    100003 2 udp 2049 nfs
    100003 3 udp 2049 nfs
    100003 4 udp 2049 nfs
    100003 2 tcp 2049 nfs
    100003 3 tcp 2049 nfs
    100003 4 tcp 2049 nfs
    100021 1 udp 32910 nlockmgr
    100021 3 udp 32910 nlockmgr
    100021 4 udp 32910 nlockmgr
    100021 1 tcp 32769 nlockmgr
    100021 3 tcp 32769 nlockmgr
    100021 4 tcp 32769 nlockmgr
    100005 1 udp 698 mountd
    100005 1 tcp 701 mountd
    100005 2 udp 698 mountd
    100005 2 tcp 701 mountd
    100005 3 udp 698 mountd
    100005 3 tcp 701 mountd
    100004 2 udp 1006 ypserv
    100004 1 udp 1006 ypserv
    100004 2 tcp 1009 ypserv
    100004 1 tcp 1009 ypserv
    100007 2 udp 889 ypbind
    100007 1 udp 889 ypbind
    100007 2 tcp 892 ypbind
    100007 1 tcp 892 ypbind

    For the client:
    rpcinfo -p mfe
    program vers proto port
    100000 2 tcp 111 portmapper
    100000 2 udp 111 portmapper
    100021 1 udp 32768 nlockmgr
    100021 3 udp 32768 nlockmgr
    100021 4 udp 32768 nlockmgr
    100024 1 udp 32768 status
    100021 1 tcp 32768 nlockmgr
    100021 3 tcp 32768 nlockmgr
    100021 4 tcp 32768 nlockmgr
    100024 1 tcp 32768 status
    100007 2 udp 1013 ypbind
    100007 1 udp 1013 ypbind
    100007 2 tcp 1017 ypbind
    100007 1 tcp 1017 ypbind

    no firewall running on either side to interfere (this checked with
    iptables -L). I see a lot of reference material to the "pre-kernel"
    nfs, but a lot to the kernel version. Hosts are fully qualified in
    /etc/hosts and the respective fstab.

    So is the nlockmgr just like rpc.lockd ? What am I missing here ?

    tia,

    -Peter


  2. Re: nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped...

    > Seeing a lot of
    > nsm_mon_unmon: rpc failed, status=-13
    > kernel: lockd: cannot monitor 192.168.1.202
    > where 202 is the client


    The NSM daemon (statd, aka status), RPC program 100024,
    is not running on the server. Regardless of whether your
    system is supposed to be running the in-kernel version
    (CONFIG_STATD) or the user-space version (/usr/sbin/rpc.statd),
    this explains your problem.

    The NSM protocol has two parts. The "local" part is between
    lockd and statd on the same machine, and that's the part which
    nsm_mon_unmon is complaining doesn't work. The "peer" part sends
    the "I have crash/rebooted" messages between clients and servers.

    Stu Friedberg

  3. Re: nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped...

    Stuart Friedberg wrote:
    > The NSM daemon (statd, aka status), RPC program 100024,
    > is not running on the server. Regardless of whether your
    > system is supposed to be running the in-kernel version
    > (CONFIG_STATD) or the user-space version (/usr/sbin/rpc.statd),
    > this explains your problem.


    So are the user-space and kernel versions interchangable ? I will go
    back and look and the .config for the kernel build.

    > The NSM protocol has two parts. The "local" part is between
    > lockd and statd on the same machine, and that's the part which
    > nsm_mon_unmon is complaining doesn't work. The "peer" part sends
    > the "I have crash/rebooted" messages between clients and servers.
    >
    > Stu Friedberg

    Thanks for the info. If anyone has a good reference on all of this, I
    would be interested in hearing about it.

    -P


  4. Re: nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped...

    On Wed, 01 Jun 2005 04:12:09 -0700, wrote:
    > So are the user-space and kernel versions interchangable ?


    Essentially. The in-kernel version introduced in Linux 2.6 (if
    I recall correctly) is an extremely stripped down statd, but it
    should be sufficient for the task.

    Stu Friedberg

  5. Re: nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped...

    Interesting... there is no CONFIG_STATD in 2.6.8.1 or later. Has this
    been absorbed into something else ?


  6. Re: nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped...

    On Thu, 02 Jun 2005 04:58:27 -0700, wrote:
    > Interesting... there is no CONFIG_STATD in 2.6.8.1 or later. Has this
    > been absorbed into something else ?


    I have been working with 2.6.5 (SLES9), and really haven't looked
    ahead. It's possible the maintainers decided it wasn't practical
    to expand on the "extremely stripped down" statd inside the kernel.
    For better or worse, Linux is a very fluid operating system from
    release to release.

    So, if there is no in-kernel statd option for your system, you need
    to find out why the rpc.statd is either not being launched or is
    failing after launch. Assuming your system startup scripts use the
    same model as SuSE's, /etc/init.d/nfslock should turn statd on whenever
    lockd is turned on, and there may be an /etc/init.d/rstatd as well,
    although the latter might not be linked as an /etc/init.d/rcX.d entry.
    rpc.statd is usually somewhat chatty, so you can probably find some
    startup messages in /var/log/messages, and perhaps a message indicating
    a failure-related problem.

    Stu Friedberg

  7. Re: nsm_mon_unmon: rpc failed, status=-13 on SuSE 9.1 - I am stumped...



    Stuart Friedberg wrote:
    > On Thu, 02 Jun 2005 04:58:27 -0700, wrote:
    > > Interesting... there is no CONFIG_STATD in 2.6.8.1 or later. Has this
    > > been absorbed into something else ?

    >
    > I have been working with 2.6.5 (SLES9), and really haven't looked
    > ahead. It's possible the maintainers decided it wasn't practical
    > to expand on the "extremely stripped down" statd inside the kernel.
    > For better or worse, Linux is a very fluid operating system from
    > release to release.
    >
    > So, if there is no in-kernel statd option for your system, you need
    > to find out why the rpc.statd is either not being launched or is
    > failing after launch. Assuming your system startup scripts use the
    > same model as SuSE's, /etc/init.d/nfslock should turn statd on whenever
    > lockd is turned on, and there may be an /etc/init.d/rstatd as well,
    > although the latter might not be linked as an /etc/init.d/rcX.d entry.
    > rpc.statd is usually somewhat chatty, so you can probably find some
    > startup messages in /var/log/messages, and perhaps a message indicating
    > a failure-related problem.
    >
    > Stu Friedberg


    Solution found... for now. I grabbed the source for the latest
    nfs-utils (which does not completely compile due to changes in
    quota.h). But at least it gets far enough that statd compiles. After
    starting that as root, all of my locking errors disappear. I also see
    RPC program 100024 show up.

    There is no mention of statd or rstatd in any of my startup scripts.
    Oddly enough it looks like it is just plain missing. I will most likely
    modify the nfslock to start and stop statd. I have looked through
    /var/log/messages severyl times, and I do not see any errors there.
    Very strange stuff indeed.

    regards,

    -Peter


+ Reply to Thread