UNSOLVED

jlcasaux

updated

18 years ago

J

jlcasaux

29 Posts

0

2453

May 8th, 2009 09:00

Message about incorrect client

Hi everybody. I have a message when trying to backup a client. The backup is successful but appears the message:

Successful Save Sets --- * sun8d7siba1:Probe savegrp: suppressed 11 lines of output. * sun8d7siba1:Probe 7198:savefs: path /informix/QAS by default belongs to client sun8d7siba1 and NOT client sun8d7siba1! * sun8d7siba1:Probe 7200:savefs: Default client index for scheduled save will be that of sun8d7siba1. * sun8d7siba1:Probe 7198:savefs: path /var/mqm by default belongs to client sun8d7siba1 and NOT client sun8d7siba1! * sun8d7siba1:Probe 7200:savefs: Default client index for scheduled save will be that of sun8d7siba1. * sun8d7siba1:Probe 7198:savefs: path /export/home by default belongs to client sun8d7siba1 and NOT client sun8d7siba1! * sun8d7siba1:Probe 7200:savefs: Default client index for scheduled save will be that of sun8d7siba1. * sun8d7siba1:Probe 7198:savefs: path /EAGENT by default belongs to client sun8d7siba1 and NOT client sun8d7siba1! * sun8d7siba1:Probe 7200:savefs: Default client index for scheduled save will be that of sun8d7siba1. * sun8d7siba1:Probe savefs sun8d7siba1: succeeded.

As you see, it seems to be a cluster issue, but the client is a single node. And the strange is that the message makes reference to the same client (sun8d7siba1).
Have you any idea of what is going on?
Thank you in advance
  • amediratta

    2 Intern

    •

    2014 Posts

    747

    1

    Posted May 10th, 2009 04:00

    It could be a mix up with the client IDs in the database somewhere. Is there a backup command given? Try and use -c sun8siba1 option in the backup command and check.
  • atremols

    29 Posts

    747

    1

    Posted May 11th, 2009 12:00

    I have seen this message if:

    - if the primary of the client is NOT the one defined in networker, check hostname output on the client and check to what is defined in networker as the client name
    - Miss-configured networker in clusters

    HTH
    Abdon
  • ble1

    6 Operator

    •

    14354 Posts

    •

    56186 Points

    747

    1

    Posted May 12th, 2009 02:00

    To add to what has been said before, I have seen this also when DNS/hosts on server/sn and client were giving different outputs.
  • DavidHampson

    2 Intern

    •

    1080 Posts

    747

    0

    Posted May 12th, 2009 04:00

    Is it correct from your message that this is not on a cluster? Could you tell us how it is configured for backups (whether you use the All saveset for example), as well as anything that is unusual about the O/S

    This appears to be an error with the probe which is what Networker runs first of all to communicate with the client and confirm the list of savesets it wishes to backup. You can run this without doing a backup by using the -p switch with savegrp from the command line - could you possibly run this command and see if it gives more information:

    savegrp -vvvv -p -c sun8d7siba1 groupname
  • jheemskerk1

    1 Message

    747

    0

    Posted May 14th, 2009 14:00

    We have same problem after upgrade from 7.2.2 to 7.4.4.
    First also thought of all the mentioned but could not find solution yet . We see it in cluster and non cluster on Solaris and HP-UX running 7.4.4. storage nodes.
  • rjgold

    21 Posts

    747

    0

    Posted May 27th, 2009 11:00

    We also get this error since upgrading to 7.4.4. I have a case open with EMC and I also saw on a NetWorker discussion list that others have cases open. It appears that the backups are successful even thought it gets this "nonsense" error message. My clients that are receiving this error have ip(s) that are on a private network and a public ip address registered in DNS. The NetWorker Client is defined with the address in DNS, with aliases field containing the private network names. Apparently, NetWorker client uses the private name in the /nsr/res/nsrladb ... thus the conflict. This is all speculation on my part based on looking at output from "nsradmin -s -p nsrexec". I heard this error message does not come out in 7.5.1 ...
  • jlcasaux

    29 Posts

    747

    0

    Posted December 14th, 2009 09:00

    Well, this problem started after migrating to 7.4.4 version. And also it happened to a lot of clients. So, we can think of make the migration soon. And is also true that the names of the clients are not the same as the hostnames, because the customer wants to make a difference between the data network and the backup network, so he uses a different name instead of the hostname to define the clients name.
  • ble1

    6 Operator

    •

    14354 Posts

    •

    56186 Points

    747

    0

    Posted December 19th, 2009 08:00

    I got this first time on 7.4.5 on HPUX (before I was running 7.4.4.5 without problems).   I've been given the patch so if you find yourself affected open the case... (in my case with standalone clients I could ignore it, but it also cause failures with cluster clients so....).