cancel
Showing results for 
Search instead for 
Did you mean: 

Lots of backups failing with status 249 or 42

lu
Level 6
Hi ! On NBU 6.5.4, we sometimes have lots of backups failing during peak backup hours with status 249 or 42. Maybe there's a congestion somewhere, but where ? I've also created the file /usr/openv/netbackup/FFREADtries and see next day if there are improvements. More ideas ? BR, Lu
4 REPLIES 4

NBU_TSE
Level 4
Employee
As the problem only happens sometimes, most intermittent problems will relate to a network problem, also you are seeing a network read failed status 42 error.

That touch file might help, but I would make sure your NIC configuration is setup so that you do not get packet drops

Make sure NIC drivers are up to date, the speed and duplex are hard coded to full duplex and same speed across each network device in the backup route.

Hope this helps

Best Regards
Gary

varunprakash
Level 5
For Error code 249:

This is occured basically  when the server timed out while waiting for the client to finish sending the file list.

The timing issue involves contention for the bpdbm process between bpbrm and bptm.  The result of this contention is that bpdbm is processing a long request from bpbrm and the request made by bptm times out before it is fulfilled.

The solution to this is to create a touch file on the master server to adjust the timeout threshold.  By default, the timeout parameters are 5 retries with a sleep interval of 60 seconds.  This default can be adjusted with the introduction of a touch file called "FFREADtries".

The location of the FFREADtries file for:
UNIX:
/usr/openv/netbackup/
 
Windows:
..\veritas\netbackup\.

Once the file has been created, values can be added to adjust the default settings.  For instance 20 60 can be added to make it have 20 tries with a sleep interval of 60 seconds.

Note:

A cycling of the Master Server services may be necessary to activate the timeout threshold introduced by the FFREADtries file.

FYI:

http://seer.entsupport.symantec.com/docs/285252.htm


For Error code 42:

All client backups using a particular media server fail, and exit with a NetBackup Status Code 42, network read failed\

Enable the bptm log file on the media server by creating the folder /usr/openv/netbackup/logs/bptm on UNIX media servers and <install_path>\VERITAS\NetBackup\logs\bptm on Windows media servers.

To resolve this error, add the media server to the master server's media server list. Check the media server and verify the media server name is there on it's own server list.

To know how to add the media server's media server list, follow the procedure in the TN given below.

http://seer.entsupport.symantec.com/docs/251899.htm


 

lu
Level 6
In fact this problem seems to be related to unexpected Solaris cluster switches. I can see the switches in /usr/openv/netbackup/bin/cluster/AGENT_DEBUG.log but the exact cause is not clear. Is there a log somewhere else which explains why the cluster switched ?

varunprakash
Level 5
To find out why the cluster failed over to passive node, check the clustee logs or OS logs.