I have run into socket write failed errors (24) a few different times.
Every time that I have, it has been related to a network issue. The funny
thing about this error is that you can often ping, ftp, etc. between the
server and the client nodes, and everything seems to work fine. For
troubleshooting I would usually clear the network statistics on the server,
then do a large ftp and look at collisions, network errors, etc. using
netstat. In a couple of cases I found out that the port that the server,
not the client, was connected to on the switch was bad and it had to be
changed. Another more likely scenario is that your network settings may be
off between your switch or hub and between your server. You want to check
to see what speed your network card is set to in your server and what it is
set to on the switch/hub. Sometimes you may have auto-negotiate, and if
this is the case, you will need to verify what the system and the switch
actually negotiated to. I like to hard set the server's network speed and
let the switch be auto-negotiate, or hard set in both places.
The last error message about the vmd not running means that your media
manager daemon, vmd, is not running correctly. You may want to stop and
restart your NetBackup processes on your media server and then use bpps -a
or a ps to verify that vmd is still running. If it isn't, then you have a
problem with your media media manager configuration. If you keep all of
your database information on your master server, then you should check to
see what size the volDB on your media manager is. I believe it should
either be 0 bytes or sometimes it might be 8 bytes (approx), but if it is
larger this may be part of your issue with vmd not running correctly.
Hope this helps.
Dan
-----Original Message-----
From: Jason Ahrens [mailto:AhrensJ AT psi DOT ca]
Sent: Monday, July 09, 2001 8:46 AM
To: Veritas BU
Subject: [Veritas-bu] Duplication failed
I had a dupication fail this weekend.. First time I've ever had this happen.
The duplicate is going fromt and to tape drives in the same silo.
02:29:21 INF - Duplicating class Unix, schedule Full
(veritasmedia1_0994570466) created 07/08/2001 01:34:26, media id 010370.
02:29:23 INF - Waiting for positioning of media id 010554 on server
veritasmedia1.
02:29:24 INF - Waiting for mount of media id 010370 on server veritasmedia1.
02:29:25 INF - Beginning duplication on server veritasmedia1 of client
veritasmedia1 image, creating copy 2.
02:30:15 INF - Waiting for positioning of media id 010370 on server
veritasmedia1.
02:30:27 INF - Beginning duplicate on server veritasmedia1 of client
veritasmedia1.
02:37:48 INF - host veritasmedia1 backup id veritasmedia1_0994570466 read
failed, socket write failed (24).
02:37:49 INF - host veritasmedia1 backupid veritasmedia1_0994570466 write
failed, Media Manager volume daemon (vmd) is not active (81).
02:37:51 INF - Duplicate of backup id veritasmedia1_0994570466 failed, Media
Manager volume daemon (vmd) is not active (81).
Any clues on what may have caused this failure?
Thanks
Jason
--
Jason Ahrens
Systems Administrator/Backup Specialist
PSINet Limited
http://www.psi.ca
The Internet SuperCarrier
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
|
