Connection problem?

Aebi

ADSM.ORG Member
Joined
Feb 8, 2005
Messages
8
Reaction score
0
Points
0
Location
Austria/Vorarlberg
Website
www.vvb.at
PREDATAR Control23

I've tried different Client-Versions (5.30C, 5.22 and 5.11C) and I always have the same problem. The OS is Win 2003 Server Standard.

But i don't know why. I've configured the Client similar to other (running) clients.

He always tries to back up something and also backs it up, but often he fails before the file is finished (filesize 800 to 900MB). Last night he backed up 2,86GB but effective he only got one file with 900MB....



So here is an excerpt from the dsmerror.log:



.....(loads of error-messages like them beneath for different files)

11.11.2005 05:21:29 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:21:29 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:21:44 ANS1810E TSM session has been reestablished.

11.11.2005 05:21:44 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:21:44 ANS1228E Sending of object '\a0100094e$MSSQLBACKUPlsVVB_db_200510182200.Bkp' failed

11.11.2005 05:21:44 Return code 0 unknown

11.11.2005 05:21:44 Unknown system error

Please check the TSM Error Log for any additional information



11.11.2005 05:21:44 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:21:59 ANS1810E TSM session has been reestablished.

11.11.2005 05:56:14 ANS1005E TCP/IP read error on socket = 880, errno = 10054, reason : 'An existing connection was forcibly closed by the remote host.'.

11.11.2005 05:56:14 cuConfirm: Received rc: -50 trying to receive ConfirmResp verb

11.11.2005 05:56:14 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:56:14 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:56:29 ANS1810E TSM session has been reestablished.

11.11.2005 06:08:59 ANS1005E TCP/IP read error on socket = 568, errno = 10054, reason : 'An existing connection was forcibly closed by the remote host.'.

11.11.2005 06:08:59 cuConfirm: Received rc: -50 trying to receive ConfirmResp verb

11.11.2005 06:08:59 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 06:08:59 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 06:09:14 ANS1810E TSM session has been reestablished.

11.11.2005 06:21:44 ANS1005E TCP/IP read error on socket = 576, errno = 10054, reason : 'An existing connection was forcibly closed by the remote host.'.

11.11.2005 06:30:17 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 06:30:18 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 06:30:32 ANS1811S TSM session could not be reestablished.

11.11.2005 06:30:33 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 06:30:34 ANS1369E Session Rejected: The session was canceled by the server administrator.



11.11.2005 06:30:34 ANS1512E Scheduled event 'INCREMENTAL_DAILY_2300' failed. Return code = 12.







Here is what dsmerror.log says:



..... (loads of error-messages like them beneath for different files)

11.11.2005 05:10:44 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:10:59 ... successful

11.11.2005 05:21:29 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:21:44 ... successful

11.11.2005 05:21:44 ANS1228E Sending of object '\a0100094e$MSSQLBACKUPlsVVB_db_200510182200.Bkp' failed

11.11.2005 05:21:44 Unknown system error

Please check the TSM Error Log for any additional information



11.11.2005 05:21:44 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:21:59 ... successful

11.11.2005 05:56:14 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 05:56:29 ... successful

11.11.2005 06:08:59 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 06:09:14 ... successful

11.11.2005 06:21:44 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 06:21:59 ... successful

11.11.2005 06:30:18 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 06:30:33 ... failed

11.11.2005 06:30:33 ANS1809W Session is lost; initializing session reopen procedure.

11.11.2005 06:30:34 --- SCHEDULEREC STATUS BEGIN

11.11.2005 06:30:34 Session established with server A01000DA: Windows

11.11.2005 06:30:34 Server Version 5, Release 2, Level 2.3

11.11.2005 06:30:34 Server date/time: 11.11.2005 06:30:34 Last access: 11.11.2005 06:30:32



11.11.2005 06:30:34 Total number of objects inspected: 11.562

11.11.2005 06:30:34 Total number of objects backed up: 93

11.11.2005 06:30:34 Total number of objects updated: 2

11.11.2005 06:30:34 Total number of objects rebound: 0

11.11.2005 06:30:34 Total number of objects deleted: 0

11.11.2005 06:30:34 Total number of objects expired: 1

11.11.2005 06:30:34 Total number of objects failed: 5

11.11.2005 06:30:34 Total number of bytes transferred: 2,86 GB

11.11.2005 06:30:34 Data transfer time: 21.567,54 sec

11.11.2005 06:30:34 Network data transfer rate: 139,37 KB/sec

11.11.2005 06:30:34 Aggregate data transfer rate: 111,20 KB/sec

11.11.2005 06:30:34 Objects compressed by: 54%

11.11.2005 06:30:34 Elapsed processing time: 07:30:31

11.11.2005 06:30:34 --- SCHEDULEREC STATUS END

11.11.2005 06:30:34 ANS1369E Session Rejected: The session was canceled by the server administrator.





Thanks if you read all this and special thanks if you have a solution. :)
 
PREDATAR Control23

You did not say whether this is Windows or Unix. I'll take a stab at one suggestion. This probably won't fix it. But, this one is always worth a look-see.



How is the nic configured on the server?

How are the ports configured on the switch?



We found out years ago that setting the nic and/or the ports on the switch to AUTO detect causes various anomalies. Typically, all of the switches are hard coded to Full/100 or Full/1000, whatever. Then, the Windows team does a software upgrade on the nic and the nic gets reset to AUTO detect. That always causes us problems.



From our experience... hard coding the nics and switches solves a lot of problems.



Paul
 
PREDATAR Control23

Aebi

While you are researching your networking issues, = here is a suggestion on the relationships between your client and your server.

First - take a look at your entire backup window. Extend it a couple of hours to insure your window of opportunity is big enough.

Second - take a look at your dsm.opt files. You appear to be "polling". What you could do for one time instance, is change from "polling" to "prompted". Let your TSM server contact the client and see what happens.

Third - extend any and all network timeouts - your messages about being forcefully closed is a sign that all sessions are being terminated just before 6am.

Lastly - this is a out on a wim - but disable WMI service as a presched, run your backup, and in your post sched, re-enable it. I mention this because there know issues with the WMI service.
 
PREDATAR Control23

sqabriel62 and stroke4, thanks for your help



Now it's working fine.

On the Server i changed the LLADRESS from 1581 to 1501. Mhmm, i can't remember entering 1581 as LLADRESS... :rolleyes: :grin:

The client was already prompting, so i didn't change this.

I also configured the NIC from AUTO to 100/Full.



There's only one thing left. Everytime i log on the client via the web-client i get tons of this in the dsmerror.log:



16.11.2005 11:29:01 Error -50 sending ht request

16.11.2005 11:29:01 Error -50 sending ht request

16.11.2005 11:29:01 Error writing to http socket.

16.11.2005 11:29:01 Error writing to http socket.

16.11.2005 11:29:01 Error -50 sending ht request

16.11.2005 11:29:01 Error writing to http socket.
 
PREDATAR Control23

Go back to using the default port number of 1500 for the client; 1581 for the web interface, insure you have the Java virtual machine daemons running. If you are firewalled, enable the port numbers.

Validate your DNS entries as well.



Steven
 
PREDATAR Control23

Do you ever say something and in the back of your mind you think of some caveats... it's just that they don't come up until you see what you said actually being done???



Changing the nic to full/100 versus AUTO... some caveats:



1) Check with your WAN team to confirm the setting on the router. The two need to match. If they don't then please coordinate with each other to set it correctly.



2) Don't make the change remotely... ie, via remote desktops or PCAnywhere (Heaven forbid). If the server doesn't like being changed on the fly or if you didnt' follow step #1 and the server quits talking to the router/switch then you won't be able to connect remotely to fix it.



3) Don't do it during production hours, unless you feel lucky and you don't have people breathing down your neck when the server hiccups or quits talking.



Where I work people are paranoid about users loosing the connection to their PST for a split second... availability... availability... availability...



I'm glad your backup is working better now.



Paul
 
PREDATAR Control23

i was facing similar issue with one of my tsm client.i re-booted the tsm client machine and everything seems fine now.

the client machine was having O/s WinNT and Exchange server.both OS and TDP backup was failing with same error mention above.
 
Last edited:
Top