BUT isn't this where the server COMMTIMEOUT is exceeded and the server
drops the session with the client???
A while back we were having similar (if not the exact thing) but all
V3 code.... and by taking the COMMTIMEOUT up to 1800 has cured it.
Yes that is a catch 22 'cause it will have to wait that long to drop
something that really died but....
later,
Dwight
______________________________ Reply Separator _________________________________
Subject: Re: tcpip failure interrupts backup @ approx. same # of fi
Author: dthorneycroft (dthorneycroft AT LACSD DOT ORG) at unix,mime
Date: 1/6/99 11:32 AM
Eric: You might want to search the archives, this problem comes
up often, and can have a lot of causes. Start by looking at
your dsmerror.log for more clues. I seem to remember a post by
Andy Raibeck that discussed several possible causes and fixes. This was
probably around mid year.
Eric LEWIS wrote:
>
> Sun OS Client version is: Version 2, Release 1, Level 0.7
> MVS Server Version 3, Release 1, Level 1.3
>
> No new client filesystems were added or changed. In fact no changes to adsm o
r
> the operation system occurred lately. It is a unix development box. Recently
a
> job has gone out of control and placed about ~20,000 small files in a backed u
p
> directory. This should not cause adsm problems.
>
> An apparent tcpip failure is interrupting a backup at approximately the same
> place most nights. Is there a possibility that this problem is related to
> "small file aggregation" on the version 3 server? (with a ver 2 client?) Can
> anyone offer any suggestions?
>
> Here is a 5 day history:
>
> Last night failed 3 times at 105,000
> Previous failed 3 times at 103,500
> Previous failed 3 times at 103,500
> Previous Successful
> Previous failed 3 times at 103,500
>
> Schedule Log
> 01/05/99 05:02:23 ANS4102I ***** Processed 103,000 files *****
> 01/05/99 05:02:24 ANS4102I ***** Processed 103,500 files *****
> 01/05/99 05:02:24 ANS4102I ***** Processed 104,000 files *****
> 01/05/99 05:02:25 ANS4102I ***** Processed 104,500 files *****
> 01/05/99 05:39:43 ANS4102I ***** Processed 105,000 files *****
> 01/05/99 05:39:43 ANS4017E Session rejected: TCP/IP connection failure
> 01/05/99 05:39:43 Total number of objects inspected: 105,210
> 01/05/99 05:39:43 Total number of objects backed up: 6
> 01/05/99 05:39:43 Total number of objects updated: 0
> 01/05/99 05:39:43 Total number of objects rebound: 0
> 01/05/99 05:39:43 Total number of objects deleted: 0
> 01/05/99 05:39:43 Total number of objects failed: 0
> 01/05/99 05:39:44 Total number of bytes transferred: 2,656.4 KB
> 01/05/99 05:39:44 Data transfer time: 0.06 sec
> 01/05/99 05:39:44 Data transfer rate: 39,430.18 KB/sec
> 01/05/99 05:39:44 Average file size: 2,526.9 KB
> 01/05/99 05:39:44 Compression percent reduction: 83%
> 01/05/99 05:39:44 Elapsed processing time:
> 00:45:38
> 01/05/99 05:39:44 ANS4847E Scheduled event 'SUN-SOLARIS-PRODUCTION-BACKUPS'
> failed. Return code = 1.
> 01/05/99 05:39:44 Sending results for scheduled event
> 'SUN-SOLARIS-PRODUCTION-BACKUPS'.
> 01/05/99 05:39:44 Session established with server ADSM: MVS
> 01/05/99 05:39:44 Server Version 3, Release 1, Level 1.3
> 01/05/99 05:39:44 Server date/time: 01/05/99 05:41:15 Last access:
> 01/05/99 04:55:36
> 01/05/99 05:39:46 Results sent to server for scheduled event
> 'SUN-SOLARIS-PRODUCTION-BACKUPS'.
> 01/05/99 05:39:46 ANS4483I Schedule log pruning started. This is the adsm
> error log from the same night. >
> 01/05/99 05:39:43 TcpRead: Zero byte buffer read.
> 01/05/99 05:39:43 sessRecvVerb: Error -50 from call to readRtn'.
> 01/05/99 05:39:43 ANS4017E Session rejected: TCP/IP connection failure
> 01/05/99 05:39:44 ANS4847E Scheduled event 'SUN-SOLARIS-PRODUCTION-BACKUPS'
> failed. Return code = 1.
>
> Following is the message text:"ANS4017E Session rejected: TCP/IP connection
> failure
>
> Explanation: An attempt to connect to the server using TCP/IP communications
> failed. This error can occur if the LAN connection went down or if your system
> administrator canceled a backup operation.
>
> System Action: Session rejected. Processing stopped.
>
> User Response: Retry the operation, or wait until the server comes back up an
d
> retry the operation. If the problem continues, see your system administrator
> for further help"
|