ADSM-L

Re: Inspecting over 200,000 objects with WinNT2000 client (takes 8 ho urs)

2002-03-08 12:55:35
Subject: Re: Inspecting over 200,000 objects with WinNT2000 client (takes 8 ho urs)
From: Daniel Sparrman <daniel.sparrman AT EXIST DOT SE>
Date: Fri, 8 Mar 2002 18:53:08 +0100

Hi Robert

First of all, the TCP/IP connection failure is due to the client stopping the communication with the server during inspection. After 15 mins, the server will automatically close the connection to the client, if no communication has occured.

Second of all, the easiest way to solve you're problem is to enable "journal based backup". This feature was implemented in version 4.2, and means that there is a separate service running on the machine, that in realtime scans which files are changed or created, and logs theese files to a journal log. When a scheduled backup is to occur, the client won't have to inspect all the files on the client, only read the journal log.

This means you will get rid of about 11.000 seconds, according to your schedlog.

When you have enabled the journal service, the next backup will be a standard incremental backup, and after that, all backups will be incremental, but journal-based. Therefore, don't think anything is wrong if the incremental backup takes a long time directly after you have enabled journal-based backup. This is normal.

Best Regards

Daniel Sparrman

-----------------------------------
Daniel Sparrman
Exist i Stockholm AB
Bergkällavägen 31D
192 79 SOLLENTUNA
Växel: 08 - 754 98 00
Mobil: 070 - 399 21D
192 79 SOLLENTUNA
Växel: 08 - 754 98 00
Mobil: 070 - 399 27 51

 
"Poetzsch,Robert E." <poetzsch AT ADP.UCHC DOT EDU>
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
2002-03-08 11:18 EST
Please respond to "ADSM: Dist Stor Manager"

To: ADSM-L AT VM.MARIST DOT EDU
cc:
bcc:
Subject: Inspecting over 200,000 objects with WinNT2000 client (takes 8 ho urs)


Here is my recent sched.log file showing almost 8 hours to process 279,057
objects.

I'm thinking that someone has some insight on how to "Optimize" a Windows
client for more efficient nightly backups. Also, there is an error message
about TCP/IP connection failure; which I believe is related to our dsm.opt
configuration due to a "time-out" condition.

The FS4500 Client is using "Windows 2000 Server" OS and TSM 3.7.1 Client,
but we will soon upgrade to TSM version 4.2.1 to see if this helps.

Thanks,

-Bob Poetzsch
Tivoli Storage Manager - System Administrator
UCONN Health Center
Farmington, CT  06030-5205
___________________________________________________________________
03/08/2002 05:34:46 Successful incremental backup of '\\fs4500\e$'

03/08/2002 05:34:47 --- SCHEDULEREC STATUS BEGIN
03/08/2002 05:34:47 Session established with server TSM: AIX-RS/6000
03/08/2002 05:34:47   Server Version 3, Release 7, Level 1.0
03/08/2002 05:34:47   Data compression forced off by the server
03/08/2002 05:34:47   Server date/time: 03/08/2002 05:32:23  Last access:
03/07/2002 21:41:48

03/08/2002 05:34:47 Total number of objects inspected:  279,057
03/08/2002 05:34:47 Total number of objects backed up:    1,805
03/08/2002 05:34:47 Total number of objects updated:          1
03/08/2002 05:34:47 Total number of objects rebound:          0
03/08/2002 05:34:47 Total number of objects deleted:     37,519
03/08/2002 05:34:47 Total number of objects failed:           0
03/08/2002 05:34:47 Total number of bytes transferred:   818.24 MB
03/08/2002 05:34:47 Data transfer time:                   73.79 sec
03/08/2002 05:34:47 Network data transfer rate:        11,353.75 KB/sec
03/08/2002 05:34:48 Aggregate data transfer rate:         29.56 KB/sec
03/08/2002 05:34:48 Objects compressed by:                    0%
03/08/2002 05:34:48 Elapsed processing time:           07:52:15
03/08/2002 05:34:48 --- SCHEDULEREC STATUS END

03/08/2002 05:34:48 ANS1017E Session rejected: TCP/IP connection failure

<Prev in Thread] Current Thread [Next in Thread>