I have a Veritas Master Server (NB DataCenter 3.4.2 running on NT4.0 Svc Pk 6). This server manages backup of 12 other servers consisting of Win2K, NT4.0 and HPUX servers via a SAN consisting of a Silkworm FC Switch and two StorageTek 9840 tape drives.
I have not much problem backing up the Windows platform servers. But I have a recurring problem with the backup of the Oracle Database in the HPUX servers.
The oracle database backup consists of a base stream (onsite or offsite) and about 12 other streams (default policy). Sometimes, one of the stream got hung resulting in no read/write activity until it was discovered by the DataCenter Operators. This also meant that the 9840 tape drive was not released resulting in many abended backup jobs (code 213 and 196). A check of the 'Detailed Status' of the stream that hung shows that it is trying to mount a tape.
What cause the oracle database backup to hang? Can the drives be released somehow based on time-out so that subsequent backups will not be affected? Any ideas on how to fix this?
I have not much problem backing up the Windows platform servers. But I have a recurring problem with the backup of the Oracle Database in the HPUX servers.
The oracle database backup consists of a base stream (onsite or offsite) and about 12 other streams (default policy). Sometimes, one of the stream got hung resulting in no read/write activity until it was discovered by the DataCenter Operators. This also meant that the 9840 tape drive was not released resulting in many abended backup jobs (code 213 and 196). A check of the 'Detailed Status' of the stream that hung shows that it is trying to mount a tape.
What cause the oracle database backup to hang? Can the drives be released somehow based on time-out so that subsequent backups will not be affected? Any ideas on how to fix this?