Oracle backup Error

[Juraj_Griga]

Hi,

we have a problem with backup of Oracle 10g databases using NetBackup 6.5.1. Backup Server is on Solaris 10, Oracle 10g is on Windows OS. Backup starts correctly and correctly backup some backup sets, but after some time backup finished with error 25 for child jobs end main job finished with error 6. Below is output from backup session. This error is just on the production server. Before, we tried it on testing Oracle server and there it works without any problem.

14.2.2008 15:18:37 - requesting resource 3582_LTO
14.2.2008 15:18:37 - requesting resource
nbumaster.NBU_CLIENT.MAXJOBS.spskapp00001
14.2.2008 15:18:37 - requesting resource
nbumaster.NBU_POLICY.MAXJOBS.Oracle_Sobota
14.2.2008 15:18:38 - granted resource
nbumaster.NBU_CLIENT.MAXJOBS.spskapp00001
14.2.2008 15:18:38 - granted resource
nbumaster.NBU_POLICY.MAXJOBS.Oracle_Sobota
14.2.2008 15:18:38 - granted resource  1053AF
14.2.2008 15:18:38 - granted resource  LTO
14.2.2008 15:18:38 - granted resource  3582_LTO
14.2.2008 15:18:38 - estimated 0 kbytes needed
14.2.2008 15:18:38 - started process bpbrm (pid=21973)
14.2.2008 15:18:48 - Error bpbrm (pid=21979) listen for client protocol
error - couldn't accept from data socket, Interrupted system call (4)
14.2.2008 15:18:42 - connecting
14.2.2008 15:18:42 - mounted 1053AF
14.2.2008 15:18:42 - positioning 1053AF to file 172
14.2.2008 15:18:48 - positioned 1053AF; position time: 0:00:06
14.2.2008 15:19:00 - end writing
14.2.2008 15:24:41 - Error bptm (pid=21978) cannot create data socket,
Connection timed out
cannot connect on socket (25)


14.2.2008 15:25:27 - requesting resource 3582_LTO
14.2.2008 15:25:27 - requesting resource
nbumaster.NBU_CLIENT.MAXJOBS.spskapp00001
14.2.2008 15:25:27 - requesting resource
nbumaster.NBU_POLICY.MAXJOBS.Oracle_Sobota
14.2.2008 15:25:28 - granted resource
nbumaster.NBU_CLIENT.MAXJOBS.spskapp00001
14.2.2008 15:25:28 - granted resource
nbumaster.NBU_POLICY.MAXJOBS.Oracle_Sobota
14.2.2008 15:25:28 - granted resource  1053AF
14.2.2008 15:25:28 - granted resource  LTO
14.2.2008 15:25:28 - granted resource  3582_LTO
14.2.2008 15:25:28 - estimated 0 kbytes needed
14.2.2008 15:25:29 - started process bpbrm (pid=22351)
14.2.2008 15:25:39 - Error bpbrm (pid=22357) listen for client protocol
error - couldn't accept from data socket, Interrupted system call (4)
14.2.2008 15:25:32 - connecting
14.2.2008 15:25:32 - mounted 1053AF
14.2.2008 15:25:32 - positioning 1053AF to file 173
14.2.2008 15:25:39 - positioned 1053AF; position time: 0:00:07
14.2.2008 15:25:51 - end writing
14.2.2008 15:31:32 - Error bptm (pid=22356) cannot create data socket,
Connection timed out
cannot connect on socket (25)


Any ideas?

Thanks

[Derek_Black]

I'd start by looking at the RMAN log and seeing what it tells you. Sometime the NetBackup logs only tell you what NetBackup cares about. If there is an authentication problem or something with the DB, then it will usually show up there. Can't really offer an explanation with limited information such as this.

[Juraj_Griga]

Hi,
yes, I already checked RMAN log (see below). It loks like network issue, not enough open tcp sockets or something like this. I tried to decrease tcp_time_wait_interval, but it didn't help....


RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of backup plus archivelog command at 02/14/2008 15:19:20
ORA-19506: failed to create sequential file, name="bk_DMPDB_6uj8msa7_1_1", parms=""
ORA-27028: skgfqcre: sbtbackup returned error
ORA-19511: Error received from media manager layer, error text:
   VxBSACreateObject: Failed with error:
   Server Status:  Communication with the server has not been initiated or the server status has not been retrieved from the serve
 
RMAN> 

[Derek_Black]

Good stuff. Looks like something to go on.  First things first, make sure the client and media server both have an Oracle license.  Then I'd probably look the bpcd logs to see what kind of connection errors you're getting. Also a good step is to try to do a RMAN backup on its own (backup to disk) just to see if RMAN is working (has proper authentication, etc).  Is a normal OS (flat-file) backup working from the client?

[zippy]

This is an problem with the server.

Note the socket error ----->>>

14.2.2008 15:24:41 - Error bptm (pid=21978) cannot create data socket,
Connection timed out
cannot connect on socket (25)

Look at the network card on this server.

Check the speed settings between each server (client master)

Check the switch (CISCO?)

Check patches (lan patches)

JD

[Juraj_Griga]

Hi,
today I made one succesfull backup. It looks that problem was really with network card. I set "tcp_time_wait_interval" (backup server on Solaris) and "tcptimedwaitdelay" (oracle client on windows) and it finished OK. So I will see whether next backups will be successfull.

[zippy]

Cool....

[Derek_Black]

Occam's razor....