Database Software

Hello,

We are experiencing an issue with the performance of our RMAN backup processes. At this time, the backup scripts used to backup the database as well as the archivelogs, are hanging at the point where a resync catalog (either implicit or explicit) is initiated. The database in question is a node within a two-node RAC environment. We are running on AIX 5L 64-bit servers and the database is approximately 5TB in size. Oracle 10.2.0.3 is the version of the Oracle software. I have a case open with Oracle now, but I thought I would post my issue in here in case anyone else has encountered the same behaviour in RMAN. Looking at the session on the target database via TOAD, I see the following SQL statement (which is the process that is not completing):

SELECT fenum fileno, TO_NUMBER (fe.fecrc_scn) create_scn,
TO_DATE (fe.fecrc_tim,
'MM/DD/RR HH24:MI:SS',
'NLS_CALENDAR=Gregorian'
) create_time,
fe.fetsn tsnum, fn.fnnam fname, fh.fhfsz fsize, fe.febsz block_size,
TO_NUMBER (fe.feofs) offline_scn, TO_NUMBER (fe.feonc_scn)
online_scn,
TO_DATE (fe.feonc_tim,
'MM/DD/RR HH24:MI:SS',
'NLS_CALENDAR=Gregorian'
) online_time,
TO_NUMBER (fe.fecps) stop_scn,
TO_DATE (fe.festt,
'MM/DD/RR HH24:MI:SS',
'NLS_CALENDAR=Gregorian'
) stop_time,
TO_NUMBER (BITAND (fe.festa, 4096)) clean_flag,
TO_NUMBER (BITAND (fe.festa, 4)) read_enabled_flag,
TO_NUMBER (BITAND (fe.festa, 64)) missing_file_flag, fe.fefdb,
fe.ferfn rfileno,
DECODE (fe.fepax,
0, 'UNKNOWN',
65535, 'NONE',
fnaux.fnnam
) aux_fname
FROM x$kccfe fe, x$kccfn fn, x$kccfn fnaux, x$kcvfh fh
WHERE fe.fepax = fnaux.fnnum(+)
AND fn.fnfno = fe.fenum
AND fn.fnfno = fh.hxfil
AND fe.fefnh = fn.fnnum
AND fe.fedup != 0
AND fnaux.fntyp(+) = 22
AND fn.fntyp = 4
AND fn.fnnam IS NOT NULL
AND fe.fenum BETWEEN :1 AND :1
ORDER BY fe.fenum


As well, I have tried implementing suggested workarounds from MetaLink for this issue, but the solution I attempted (add missing indexes to the RSR table), made no difference.

Any assistance would be greatly appreciated!

Thanks,

Scott