08-15-2013 06:59 AM
I am haveing a recurring issue with an NDMP backup. First the environment:
Master Server: Window 2008 R2 Enterprise NetBackup 7.0.1
Media Server: Windows 2003 Enterprise Netbackkup 7.0.1
NDMP is an EMC NX4
I have 3 policies running on this for a total of 11 paths and 1 path fails with a 13 error. The others run fine. Most of the time, it's the full that fails but the incremental backup has fail before. The following are the error messages in the job detail.
8/15/2013 9:39:08 AM - Error bpbrm(pid=1608) db_FLISTsend failed: file read failed (13)
8/15/2013 9:39:09 AM - Error ndmpagent(pid=3784) terminated by parent process
8/15/2013 9:39:15 AM - Error ndmpagent(pid=3784) NDMP backup failed, path = /hosted_qa
8/15/2013 9:39:45 AM - Error ndmpagent(pid=3784) NDMP_LOG_ERROR 0 Medium error
8/15/2013 9:39:46 AM - Error ndmpagent(pid=3784) NDMP_LOG_ERROR 0 Backup is aborted.
file read failed(13)
This path could run fine for a couple of weeks then fail over and over. The storage team has looded and can't seem to find anything.
08-15-2013 08:10 AM
If backup is taking a long time try creating the NDMP_PROGRESS_TIMEOUT file on the NetBackup media server per this technote:
http://www.symantec.com/business/support/index?page=content&id=TECH21246
Another possibility, collect the log from the EMC Data Mover
http://www.symantec.com/business/support/index?page=content&id=TECH145721
08-15-2013 10:34 AM
Check the NDMP logs in the filer, there are 5 phases in NDMP, normaly phase 1 and 2 are to index the files and phase 3 writes the directory structure, phase 4 dumps the data and 5 closes the process, normaly the first 3 phases can take long time to build the index and directory based on the number of files you may have if this is the case you will probably need to split your backups in some few folders to reduce the indexing load.
Hope this helps.
regards.
08-16-2013 02:01 AM
Please enable the such logging
legacy - bpbrm logging (verbose=5)
unified
ndmpagent (originator ID 134)
ndmp (originator ID 151)
nbpem (originator ID 116)
nbjm (originator ID 117)
nbrb (originator ID 118)
run on control station server_df command - to see if this path is still mounted (if this is file system not a dir in other file system).
check the server_log command for any issues...
Retry the job and check the above logs for more clues...
08-16-2013 07:14 AM
Check the volume name and space, if they dont have space the backup will fail.
08-16-2013 07:55 AM
Here is the troubleshooting TechNote for NDMP:
http://www.symantec.com/docs/TECH56492
This looks like possibly a media or hardware error. Possibly. So you will want to collect all the logs in the above article.
09-04-2013 05:48 PM
09-05-2013 04:30 AM
hav u tried backup using SNAPSURE = Y option.
09-05-2013 04:36 AM
thread opened on 15th August - several suggestions but never an answer from the thread creator - I assume it is fixed by now!
09-05-2013 06:18 AM
09-10-2013 07:04 AM
Sorry, I'm having issues with getting others involved. I am not in the storage team and needed to wait and get someone to look into the logs. I can tell you you that it's just this path that fails and there are 11 other paths on the same unit that back up fine. This has been a challenge for me. I'm eing asked if I can see what folder in the path is getting hung. I am going to open a case with Symantec and I will post here when the issue is resolved. Sorry for the delay.