Re: rman backup slow
- From: Prasad <p4cldba@xxxxxxxxx>
- To: "Mark W. Farnham" <mwf@xxxxxxxx>
- Date: Sun, 1 Feb 2009 14:10:19 -0800
Mark,
I see the x$dual wait while rman does the backup archivelog delete all input
as part of the db backup . Here is a output from v$sql
1* SELECT * FROM V$SQL WHERE HASH_VALUE=4913485
SQL> /
SELECT
TO_CHAR(SYSDATE,:"SYS_B_0",:"SYS_B_1"),TO_CHAR(SYSDATE,:"SYS_B_2",:"SYS_B
_3"),TO_CHAR(SYSDATE,:"SYS_B_4",:"SYS_B_5") FROM X$DUAL
24562 1392 2008 0
1 0
13 13 13 0 1
2009-01-30/21:00:57 0 13 0 0
13 3 CHOOSE 0 0
0 0 00000402D5A1F648 00000402FBA79CC0 4913485
3957808902 0
backup archivelog: ORA_DISK_1 1350560595
0000024 FINISHED 1605141510
0
10000
4092 0 00000402FF203628 6 N VALID
0 2009-01-30/21:00:57 N
but I think the real culprit is disk async IO wait . because everytime I
start taking the backup the statspack shows it as the top wait .
I did a level 12 trace . and here is the last 5% backup . I see there are so
many sleeping for 10 sec wait .not very sure how to proceed from here :)
command backup:95.1% complete, time left 00:01:08
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxr: callback returned TRUE, skipping sleep
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxr: sleeping for 10 seconds
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxr: sleeping for 10 seconds
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxr: sleeping for 10 seconds
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxr: sleeping for 10 seconds
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxr: sleeping for 10 seconds
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxr: sleeping for 10 seconds
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxr: sleeping for 10 seconds
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxr: sleeping for 10 seconds
DBGRPC: krmxpoq: xc=4308834080, action="0000012 STARTED", col_l=15, ind=0,
sid=15
DBGRPC: krmxpod: xc=4308834080 has finished remote rpc #12
DBGRPC: krmxr: xc=4308834080 chid=ORA_DISK_1 calling peicnt
DBGRPC: krmxrpc: xc=4308834080 kpurpc2 rc=0 db=target
proc=DBMS_BACKUP_RESTORE.BACKUPPIECECREATE
DBGRPC: krmxrpc: xc=4308834080 finishing RPC #12 to target
DBGIO: Type %Comp Blocks Tot Blocks Blksize ElpTime(s) IO Rt(b/s)
Name
DBGIO: ---- ----- ---------- ---------- ---------- ---------- ----------
-----
DBGIO: IN 100.0 1048576 1048576 8192 1405 0
/rad/u4/oradata/casemt/casemt01.dbf
DBGIO: OUT 1040398 8192 1405 3009233
/rad/ora_bkup/rman/CASEMT_5213_1_6776
78395.dmp
DBGIO: AGG 1048576 8192 1405 0
channel ORA_DISK_1: finished piece 1 at 01-FEB-09
channel ORA_DISK_1: finished piece 1 at 01-FEB-09
piece handle=/rad/ora_bkup/rman/CASEMT_5213_1_677678395.dmp comment=NONE
channel ORA_DISK_1: backup set complete, elapsed time: 00:23:35
DBGRPC: krmxr: xc=4308834080 finished step
DBGRPC: ENTERED krmqgns on [02/01/2009 12:23:30]
DBGRPC: krmqgns: looking for work for channel default (krmqgns)
DBGRPC: krmqgns: commands remaining to be executed: (krmqgns)
DBGRPC: CMD type=backup id=1 status=STARTED
DBGRPC: 1 STEP id=1 status=FINISHED
DBGRPC: krmqgns: no work found for channel default (krmqgns)
DBGRPC: (krmqgns)
thanks
pk
Other related posts:
