Hi all,

Env: DB2 ESE/DPF 8.1.6 64 Bit on AIX 5.2.0

we recently had problems with an offline backup of a database with 5
partitions that fails with SQL0956 - Not enough storage in dbheap.
DBHEAP at that time was 5000 Pages per partition. The last backup ran
just fine with that dbheap - no objects (bufferpools, tablespaces,
tables ..) were added or changed since then. We rerun the failing
backup with greater dbheap sizes until it succeeded with a dbheap of
20000 pages. During backup we took snapshots to watch the dbheap high
watermark. We saw that dbheap grows slowly but at the end of the backup
(last 2 to 3 minutes) it grew dramatically. I know that at the end of
backup the recovery history file gets pruned and the log control file
is put into the backup. Can this be the reason for such more dbheap
requirement ?
Can someone share some insights into usage of dbheap during backup ?

the backup ran like this:
db2 backup db <dbname> use tsm compress

TIA for any answer

Joachim