Suzanne,
It sounds as if something is causing network connections to time out. If you
have multiple routes to the hosts it is possible that with a network load,
the traffic attempted to route a different way which resulted in a failure.
You can inspect your network and see if there are any DNS issues, VPN
issues, routing issues or if a required_interface selection in NetBackup may
be warranted.
It sounded like you had only one media server. If you have another media
server other than the Master/Media you listed, make sure that they are at
the same patch level.
Review the bpbrm,bpdbm, and bprd logs on the master and the bpcd and bpbkar
logs on the client.
Skip
-----Original Message-----
From: Suzanne Palmer [mailto:s.palmer AT umassp DOT edu]
Sent: Monday, November 04, 2002 11:05 AM
To: Veritas-Bu
Subject: [Veritas-bu] error 25
Okay, I'm stumped. I'm getting intermittent failures with error code
25. In some cases a job will run for a considerable amount of time
before failing, then the job is requeued. Sometimes it then runs
successfully, sometimes not.
Here are the relevant log lines (this is after running for more than
40 min):
03:54:42 [23063] <4> getsockconnected: gethostbyname failed for
nbserver.umassp.edu: HOST_NOT_FOUND (1)
03:54:42 [23063] <16> getsockconnected: cannot find host:
nbserver.umassp.edu, h_errno=1
03:54:42 [23063] <16> db_begin: getsockconnected() failed: No route to
host (148)
03:54:42 [23063] <16> db_FLISTsend: db_begin() failed: cannot connect
on socket
03:54:42 [23063] <16> bpbrm main: db_FLISTsend failed: cannot connect
on socket (25)
03:54:52 [23063] <4> getsockconnected: gethostbyname failed for
nbserver.umassp.edu: HOST_NOT_FOUND (1)
03:54:52 [23063] <16> getsockconnected: cannot find host:
nbserver.umassp.edu, h_errno=1
03:54:52 [23063] <16> db_begin: getsockconnected() failed: No route to
host (148)
03:54:52 [23063] <16> db_ERROR: db_begin() failed: cannot connect on
socket
03:54:52 [23063] <16> log: db_ERROR() failed: cannot connect on socket
(25)
03:56:24 [23063] <4> bpbrm Exit: client backup EXIT STATUS 25: cannot
connect on socket
This behavior is wildly inconsistent; I have 22 hosts I currently back
up, and some of them fail this way each and every time. Some fail some
of the time. Some never fail. The hosts that fail most often are
larger, but I don't know if that's relevant. Some are windows
machines, some are Unix, some are on the same subnet as the server,
some not. I wish I had more concrete information, but I have been
unable to find any common element. )-:
I'm running the 3.4 server on a solaris 8 system. I tried updating the
install with the latest patches from veritas, but that had no visible
effect.
Thanks in advance!
-Suzanne
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
|
