Netbackup are running for too long time

[hugo_lkp]

I am using Netbackup 6.5 to backup my database with RMAN. I am using their script hot_tablespace_backup.sh to backup
to tap drive. I got the problem related the elapsed time is 2 hours to backup a 300G big datafile...


*******************************************************************************************************************
tail -f hot_tablespace_backup.sh.out

allocated channel: ch01
channel ch01: SID=124 device type=SBT_TAPE
channel ch01: Veritas NetBackup for Oracle - Release 6.5 (2009050106)

Starting backup at 28-AUG-09
channel ch00: starting incremental level 0 datafile backup set
channel ch00: specifying datafile(s) in backup set
input datafile file number=00006 name=+BIG_DG/test/datafile/test_tablespace.269.694718769
channel ch00: starting piece 1 at 28-AUG-09
...


*******************************************************************************************************************


On the other hand, I have testing RMAN to backup my database without Netbackup...Only backup to local disk...also is backup a same datafile
I found that the backup running time only 10 mins .........So that, I don't know which part got the problem

[quebek]

I would check what backup speeds to you get for this oralce client with standard file system backup? Maybe there is a network issue?
Also you can try to FTP some files from this client to nbu server - to test the network throughput.

[Taqadus_Rehman]

Please review this white paper covring bestpractices

http://h71036.www7.hp.com/ERC/downloads/4AA0-8102ENW.pdf

and do study

Netbackup Performance Tuning Guide
NetBackup Admin Guide for Oracle

[hugo_lkp]

I had use FTP transfer a 2GB size file ~~~ the speed around ~11.5 MB/s. It is ok or not?

[hugo_lkp]

Backup client database rman config setting :

CONFIGURE RETENTION POLICY TO REDUNDANCY 1; # default
CONFIGURE BACKUP OPTIMIZATION OFF; # default
CONFIGURE DEFAULT DEVICE TYPE TO DISK; # default
CONFIGURE CONTROLFILE AUTOBACKUP OFF; # default
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '%F'; # default
CONFIGURE DEVICE TYPE DISK PARALLELISM 1 BACKUP TYPE TO BACKUPSET; # default
CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE MAXSETSIZE TO UNLIMITED; # default
CONFIGURE ENCRYPTION FOR DATABASE OFF; # default
CONFIGURE ENCRYPTION ALGORITHM 'AES128'; # default
CONFIGURE COMPRESSION ALGORITHM 'BZIP2'; # default
CONFIGURE ARCHIVELOG DELETION POLICY TO NONE; # default
CONFIGURE SNAPSHOT CONTROLFILE NAME TO '/oracle/11.1.0/dbs/snapcf_test.f'; # default


Netbackup server config setting:

SIZE_DATA_BUFFERS= 262144
NUMBER_DATA_BUFFERS=32

It is ok?

[hugo_lkp]

Backup client database rman config setting :

CONFIGURE RETENTION POLICY TO REDUNDANCY 1; # default
CONFIGURE BACKUP OPTIMIZATION OFF; # default
CONFIGURE DEFAULT DEVICE TYPE TO DISK; # default
CONFIGURE CONTROLFILE AUTOBACKUP OFF; # default
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '%F'; # default
CONFIGURE DEVICE TYPE DISK PARALLELISM 1 BACKUP TYPE TO BACKUPSET; # default
CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE MAXSETSIZE TO UNLIMITED; # default
CONFIGURE ENCRYPTION FOR DATABASE OFF; # default
CONFIGURE ENCRYPTION ALGORITHM 'AES128'; # default
CONFIGURE COMPRESSION ALGORITHM 'BZIP2'; # default
CONFIGURE ARCHIVELOG DELETION POLICY TO NONE; # default
CONFIGURE SNAPSHOT CONTROLFILE NAME TO '/oracle/11.1.0/dbs/snapcf_test.f'; # default


Netbackup server config setting:

SIZE_DATA_BUFFERS= 262144
NUMBER_DATA_BUFFERS=32

It is ok?

[quebek]

Is it 1Gbit or 100 Mbit? I suspect it is 1Gbit. So this xfer is not so awsome :D
Lets do some maths:
300GB -> 307200 MB
1Gbit practical throughput is arouund 80 MB/s
so to xfer these 300GB over ethernet we need to spend about 1 hour
307200/80 = 3840 sec = 1,06 hour
so this is the max you can achive doing LAN backup. Your time was something about 2 hours. Not so bad :D, but still performance can be better - you need to dig in performance guide provided by Taqadus.
to have similar results to backup to disk - you should make this client an Enterprise client or Media server and backup it over SAN to tape drives directly.

[hugo_lkp]

It is 1Gbit.
actually...I have 300G data need to backup...but there is only one table one row... so that my backup file (s248_1251428080_C1_F1.1251428080.img) size only have 1312 KB....But I don't know why there are backup running for a long time.
On the other hand. You can use teamviewer remote login to review If you have time

[hugo_lkp]

there is a Bigfile tablepsace <--- This is a issue or not

[Jim_Horalek]

Post your network buffer size. NET_BUFFER_SZ,(same location as others)  If it doesn't exist your buffer is probably too small - default 32K I believe. No magic number here but probably should be at least 128K.

 

[hugo_lkp]

my NET_BUFFER_SZ is 65536