Archiving via file list takes hours until start

waelti

ADSM.ORG Member
Joined
Apr 20, 2010
Messages
54
Reaction score
2
Points
0
Location
Germany
Hello,
we updated from Spectrum Protect Version 7 to Version 8 a month ago. The server runs under Redhat 7, there are no problems on the server side. We have also updated the backup clients from version 7 to version 8 (8.1.10.0).
When archiving via file list we now have the problem that the Windows client displays the message "Archive function invoked" and in the task manager the memory consumption increases the larger the file list is, but the actual archiving only starts after several hours. Currently I have an archive over 2.2TB with over 330 thousand files and I have been waiting for about 3 hours. The client currently requires over 5 GB of RAM and is still increasing. On the server you don't see any load on the archive instance. At some point the archiving starts.
Do I have to change something in the option-file so that the client works under version 8 like it does under version 7? When I run a retrieve everything works as before.
Kind regards
Walter Reis
 

Attachments

  • dsm_arc1.opt.txt
    359 bytes · Views: 2
What's the full command that you use to kick off the archive?

Can you put the content your option file in the post? I can't speak for others, but I don't download files from forums.
 
The content of the option file is:

NODENAME IN-PCSERVER
PASSWORDACCESS GENERATE
SCHEDMODE PROMPTED
TCPSERVERADDRESS 192.168.120.102
TCPPORT 1611
ERRORLOGNAME C:\Temp\dsmerror_arc1.log
ERRORLOGRETENTION 7 D
SCHEDLOGNAME C:\Temp\dsmsched_arc1.log
SCHEDLOGRETENTION 7 D
Skipntpermissions yes
dirmc dirmc
RESOURCEUTILIZATION 10
SKIPNTSECURITYCRC YES


The full command is for archiving files:
dsmc archiv -filelist=\\in-sv-nas01\export$\Archivdaten.txt -deletefiles -description="Archive Date: 2020-10-15 Ticketnr. 123456"

In client version 8 it depends on the file size when the archiving starts after the call. With 335 thousand files about 3.5 hours, with 7800 files about 4 minutes. In version 7, archiving started after 5 to 10 seconds regardless of the file size.
 
And the only change was the upgrade from 7 to 8? Using the same command before and after? If that's the case, you should probably open a case with IBM.
 
The only change was the upgrade from 7 to 8. All other stays the same.
I opened a call at IBM and hope that it will be more professional like my last one. We had the problem back then when we were upgrading from version 5 to 7 that opening the Archive GUI suddenly took several hours instead of some seconds. It was an eternal back and forth with IBM with the logfiles and settings, but nothing came out, so I closed the call annoyed. This has not changed in version 8. The command line worked normally at that time. Now, unfortunately, the command line has also been affected. The difference between the GUI and the command line is, that the command line hardly puts any load on the database, with the GUI it runs until it opens on 100% CPU load.
 
Back
Top