10-18-2018 02:57 AM
Dear All,
My current system run netbackup schedule backup, i can see the netbackup server try to mount the media but then network time out
Solved! Go to Solution.
10-24-2018 12:03 AM
@Home_224 wrote:
Hi Marianne,
I can see the two log directory existing but there is no logging in the two directory.
Simply not possible if you have regular backups running.
Each time a Standard backup runs for this client, bpbkar process is started on the client.
If bpbkar folder exists under /usr/openv/netbackup/logs, then bpbkar will create a new file for that day and log activities.
Same for bptm and bpbrm.
These 2 processes run on the media server.
If the master is media server as well, then these processes run on the master.
If log folders exists under /usr/openv/netbackup/logs, then each bptm and bpbrm process will log to the <date>.log file.
Double check folder names - check for typo errors.
Check location of directories - they must be under /usr/openv/netbackup/logs (not /usr/openv/logs).
10-18-2018 02:58 AM
The client is run on solaris, I try to set the timeout to 300 and check the result after the following day, but it sames symptoms.
Please assist with me to give suggestion to fix it.
Many thanks
10-18-2018 03:08 AM
Hello,
can you please give us some further information regarding your configuration?
Is it a physical or virtual tape library you're writing on?
How is it connected to the media server (which protocols, cables, switches, etc.)?
Thanks.
10-18-2018 03:23 AM
Hello,
My current backup solution is quite simple, there is one master server connection with FC tape to run the backup, all the client use the backup lan for data backup. I tried to use the another standby backup server to run the same policy backup to regular running backup , the client is backup successfully.
We use the policy to run the regularaly backup for full or incremential , it is the same network error connection .
Thanks,
Alfred
10-18-2018 03:25 AM
This has nothing to do with tape trying to mount.
You can see in the the 2nd screenshot that the tape has actually mounted.
'Begin writing' means that the tape drive has mounted and positioned the tape and is ready to start writing data as soon as first data buffer is received from the client.
If no data is received with the 5 min timeout (300 sec), then we see bpbrm reporting a timeout.
The policy name suggests Oracle backup.
Is this Oracle Policy type?
Or a Standard policy type that is backing Oracle dumps/disk-based backups?
For Oracle policy type, the recommendation is to turn up Client Connect and Client Read Timeout on the media server to 1800 (30 min).
You need these logs on the media server to see what is happening between the tape mount and the timeout:
bpbrm (will show client connect, bptm started and metadata being received)
bptm (will show data being received from the client and written to tape)
You also need to check logs on the client - if Oracle policy type, check dbclient log.
If Standard policy, check bpbkar log.
All of these log folders need to be manaully created if they do not exist already.
dbclient folder needs 777 permission.
Level 3 logs should be fine - level 0 logs do not give enough info and level way more than is required for day-to-day troubleshooting.
(Veritas Support always asks for level 5 logs, but irrerelevant here as your OS and NBU versions are no longer supported by any vendor.)
PS:
I know from previous posts that you have challenges to access your client's environment, but troubleshooting this kind of issues really needs access to all relevant parties, servers and clients and log files.
There is no other way.
10-18-2018 06:43 AM - edited 10-18-2018 06:45 AM
Thank you Marianne,
We use the policy type is standard for all server, I will gather the log with end user for analysis asap
Thank you again !
10-19-2018 02:18 AM - edited 10-19-2018 02:22 AM
If you need our assistance to analyze logs, please copy the logs to .txt files, e.g.
bptm.txt
bpbrm.txt
bpbkar.txt
We need one set of full logs that contains the info that we see in Activity Monitor.
If the logs are for a different error (on another day), then please include all the text in Activity Monitor Details.
Upload the .txt's here.
10-19-2018 02:26 AM
Hi Marianne ,
Thank you very much!
I will upload three log next Monday, and thank you again
10-21-2018 08:36 PM
10-22-2018 12:04 AM
Okay, so, the screenshot shows a problem for 22 August.
Are you now, 2 months later trying to troubleshoot an issue that is no longer present?
Backups are now running successful?
The NBU defaults are to keep Job Monitor logs for 3 days, and regular logs for 28 days.
If you want to troubleshoot backup failures, best to do that ASAP.
10-22-2018 01:08 AM
I have the another one for daily backup server, but the configuration is not powerful than the fault one. So I try to get the bptm log which is in zip format.
BTW, how can i enable the two log bpkar if it for next time for troubleshooting
10-22-2018 03:38 AM
Just bptm log alone won't help.
You need to follow the entire process flow as seen in Job Details :
bptm and bpbrm on the media server
bpbkar on the client.
You enable process logging by creating the directories on the relevant server under /usr/openv/netbackup/logs.
Please double-check spelling - client log is bpbkar (not bpkar).
I don't understand what 'two log bpkar' means....
10-22-2018 07:58 PM
Hi Marianne,
I can see the two log directory existing but there is no logging in the two directory.
10-24-2018 12:03 AM
@Home_224 wrote:
Hi Marianne,
I can see the two log directory existing but there is no logging in the two directory.
Simply not possible if you have regular backups running.
Each time a Standard backup runs for this client, bpbkar process is started on the client.
If bpbkar folder exists under /usr/openv/netbackup/logs, then bpbkar will create a new file for that day and log activities.
Same for bptm and bpbrm.
These 2 processes run on the media server.
If the master is media server as well, then these processes run on the master.
If log folders exists under /usr/openv/netbackup/logs, then each bptm and bpbrm process will log to the <date>.log file.
Double check folder names - check for typo errors.
Check location of directories - they must be under /usr/openv/netbackup/logs (not /usr/openv/logs).