cancel
Showing results for 
Search instead for 
Did you mean: 

RMAN FLAT FILE BACKUP size 3TB to be taken on tape failed with error code 13.

Vinayak_Chennur
Level 4

Hi Friends,

I am taking a RMAN flat file backup of 3TB DB.

While taking it the backup get failed with error code 13 after completed almost 1 TB .

again i have to resume it manually & backup get starts same thing repeat till it complete 3TB.

also its take time to compelete around 10 to 15 hrs.

Kindly suggest me what is the best method to take 3TB Rman Flat file database backup on tape.

kindly find the logs:

 

nfo nbjm (pid=19007) starting backup job (jobid=87437) for client Client name, policy Policy name, schedule Daily-Diff 01/16/2017 12:38:21 - Info nbjm (pid=19007) requesting STANDARD_RESOURCE resources from RB for backup job (jobid=87437, request id:{9099BA2E-DBBA-11E6-8DC2-D60DBD1D772A}) 01/16/2017 12:38:21 - requesting resource STRG_GROUP 01/16/2017 12:38:21 - requesting resource Master_server.NBU_CLIENT.MAXJOBS.Client name 01/16/2017 12:38:21 - requesting resource Master_server.NBU_POLICY.MAXJOBS.Policy name 01/16/2017 12:38:21 - Waiting for scan drive stop HPMSLD2, Media server: Media server 01/16/2017 12:38:22 - granted resource Master_server.NBU_CLIENT.MAXJOBS.Client name 01/16/2017 12:38:22 - granted resource Master_server.NBU_POLICY.MAXJOBS.Policy name 01/16/2017 12:38:22 - granted resource 111111 01/16/2017 12:38:22 - granted resource HPMSLD2 01/16/2017 12:38:22 - granted resource Media server-hcart3-robot-tld-0 01/16/2017 12:38:22 - estimated 0 kbytes needed 01/16/2017 12:38:22 - Info nbjm (pid=19007) resumed backup (backupid=Client name_1484530211) job for client Client name, policy Policy name, schedule Daily-Diff on storage unit Media server-hcart3-robot-tld-0 01/16/2017 12:38:22 - started process bpbrm (pid=86677) 01/16/2017 12:38:23 - connecting 01/16/2017 12:38:23 - mounting 111111 01/16/2017 12:38:23 - Info bpbrm (pid=86677) start bpbkar on client 01/16/2017 12:38:23 - Info bpbkar (pid=45657) Backup started 01/16/2017 12:38:23 - Info bpbrm (pid=86677) Sending the file list to the client 01/16/2017 12:38:23 - connected; connect time: 0:00:00 01/16/2017 12:39:25 - Info bptm (pid=86679) media id 111111 mounted on drive index 0, drivepath /dev/nst1, drivename HPMSLD2, copy 1 01/16/2017 12:39:26 - mounted 111111; mount time: 0:01:03 01/16/2017 12:39:26 - positioning 111111 to file 2 01/16/2017 12:40:27 - positioned 111111; position time: 0:01:01 01/16/2017 12:40:27 - begin writing 01/16/2017 14:19:02 - end writing; write time: 1:38:35 file read failed (13)

1 ACCEPTED SOLUTION

Accepted Solutions

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

If you had NBU 7.7.x or later with NetBackup Appliance, I could suggest Oracle Copilot.

I have no other advice since your current method is not considered best practice. 

Just wishing you good luck! 

View solution in original post

7 REPLIES 7

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

Rman backup to disk and then flat file backup to tape is not best practice.

Rather purchase Database and Application agent and use rman to backup directly to tape.

There are lots of Oracle documetation on how to performance tune rman backups using backup application agent (e.g. NetBackup).

Hi Marianne,

 

I have already suggested to the management about this. But they are like the rman backup which they dump its take hardly some time and its complete quickly later we can take on tape..On tape the time take more due to which they are not ready for rman backup to tape.

 

Can you help me what is the other way of taking this backup.

 

regards,

Vinayak Chennuri

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified

If you had NBU 7.7.x or later with NetBackup Appliance, I could suggest Oracle Copilot.

I have no other advice since your current method is not considered best practice. 

Just wishing you good luck! 

:(

One thing I notice, is that you tape drive(s) does not seem to be using to recommend DATA_BUFFER_SIZE & NUMBER_DATA_BUFFERS for enterprise tape drives.

I usually start with

262144 and 256 for LTO3 and above

Of course this only helps if you can read the file(s) that fast. Use the bpbkar redirected to /dev/null to see that (see technote for precise command)

The standard questions: Have you checked: 1) What has changed. 2) The manual 3) If there are any tech notes or VOX posts regarding the issue

Just a thought more, being a Oracle server has the file system cache been limited or disabled on the server or mount point ?

The standard questions: Have you checked: 1) What has changed. 2) The manual 3) If there are any tech notes or VOX posts regarding the issue

increase the Client Read Timeout on the media server.  Start by doubling it from 300 to 1200, and so on until the backup is successful.

 

Also is it possible to use Flashbackup to backup the file or a block level backup on the disk as this will backup much faster. 

Aslo Can you upgrade the NIC on the media server and the RMAN FLAT FILE dump server to 10 gig? This will give you faster performace and allow the backups to complete.