rman backup hunging

Question

Hi, Environment - server: windows 2000 server - client: redflag x86_64 bit - target database: Oracle9i Enterprise Edition Release 9.2.0.4.0 - 64bit Production - catalog database : Oracle9i Enterprise Edition Release 9.2.0.1.0 - Production - Networker 7.7.1 bulid 230 - NMO4.2 The problem is when We try to execute the RMAN backup scripts, the process is hunging, and no error messages in daemon.log and msglog.log, using NSR_DEBUG_FILE parms,there also no messges in debugfile. using debug trace, the result as follow: #rman target / rcvcat rman/rman@ORCL_10.102.1.198 debug trace = /nsr/hunglog Recovery Manager: Release9.2.0.4.0 - 64bit Production Copyright (c) 1995, 2002, Oracle Corporation. All rights reserved. connect to target database: ORCL (DBID=1137566233) Connected to Recovery Catalog Database RMAN> @/nsr/scripts/orafull RMAN> #connect target rman/rman@orcl_10.102.1.200 2> #connect rcvcat rman/rman@ORCL_10.102.1.198 3> run { 4> allocate channel D1 type 'SBT_TAPE' 5> parms 'ENV=(NSR_SERVER=backup,NSR_CLIENT=oradb1,NSR_DATA_VOLUME_POOL=ora)'; 6> #allocate channel t2 type 'SBT_TAPE' 7> #parms 'ENV=(NSR_SERVER=backup,NSR_CLIENT=oradb1,NSR_DATA_VOLUME_POOL=ora)'; 8> #allocate channel t3 type 'SBT_TAPE' 9> #parms 'ENV=(NSR_SERVER=backup,NSR_CLIENT=oradb1,NSR_DATA_VOLUME_POOL=ora)'; 10> #sql 'alter system archive log current'; 11> backup full filesperset 1 12> format 'bk_s_p_t' 13> (database include current controlfile); 14> release channel D1; 15> #release channel t2; 16> #release channel t3; 17> } NetWorker: Reverse DNS lookup failed for address 10.102.1.200 NetWorker: Reverse DNS lookup failed for address 192.168.1.102 channel: D1 channel D1: sid=45 devtype=SBT_TAPE channelD1: NMO v4.2.0.0 start backup at 28-5 -07 channel D1: starting datafile full backupset channel D1: specifying datafile(s)from backup set. input datafile fno=00015 name=/oradata/orcl/GMISDATA.dbf channel D1: starting segment 1 at 28-5 -07 command finished backup:2.8, remains: 03:31:18 command finished backup:3.8, remains: 03:24:01 command finished backup:3.8, remains: 04:14:30 command finished backup:3.8, remains: 04:14:30 command finished backup:3.8, remains: 04:14:30 (hunging here) #cat /nsr/hunglog DBGRPC: krmxr: sleeping for 10 seconds DBGRPC: krmxpoq: xc=182918965648, action='0000012 STARTED', col_l=15, ind=0, sid=45 #sqlplus '/as sysdba' SQL> select * from v$session_wait where wait_time = 0 and sid = 45; SID SEQ# EVENT P1TEXT P1 P1RAW P2TEXT P2 P2RAW P3TEXT P3 P3RAW WAIT_TIME SECONDS_IN_WAIT STATE 45 4232 sbtinfo2 0 00 0 00 0 00 0 270 WAITING Do you know what the solution is?

rman2 · Answer

sorry, Networker 7.1 bulid 230,not Networker 7.7.1 bulid 230

ble1 · Answer

Ups, that is old and a bit unsupported already. I'm not sure how 4.2 is good with 7.1 (and I know 7.1 is really really bad). I would suggest to update first and then try again. On what OS is that? Also, are IPs 10.102.1.200 and 192.168.1.102 being used?

rman2 · Answer

client OS is REDFLAG ,the same like REDHAT,and 10.102.1.200 is client virtual IP, 192.168.1.102 is client heartbeat IP.

ble1 · Answer

Is it possible that you use virtual IP for communication towards server? In that case put name of it to hosts on both server and client as NW won't run unless it can resolve IP that is using.

dk3 · Answer

I assume that oradb1 is your virtual node, right? Could you try using FQDNs for your NSR* variables (eg. NSR_CLIENT=oradb1.domain.com instead of oradb1) in your RMAN script? We had the same problems some weeks ago and it seemed that RMAN cannot find the short client name in Networker when you've created clients in NW with FQDNs.

huenian · Answer

I have the same problem as yours. Now, it was solved.The solution may be:delare theses address 10.x.x.x in (addresses in error: DNS...) in your /etc/hosts file of your oracle database.

dk3 · Answer

Hi huenian,I'm having the reverse DNS problems on our site too - but RMAN backup of the database is not disordered by this.regards.

ble1 · Answer

DNS problems affect backups only if NIC with problems is the one used by configuration or/and communication path towards server or storage nodes.

otmanix · Answer

Hello,on Oracle 10 there is an issue with large database backups and RMAN because of RMAN using cost based optimizer for backup statistics. There's also a workaround by forcing the RMAN session to use rule based optimizer.I have written a post for that earlier. You only have to search for it.Perhaps thar solves your problem.Best regards, Otmar Meier

NetWorker

Was this post helpful?