-----John D. Schneider wrote: ----- >2) Another unusual circumstance is that the 'backup stgpool' for the >storage pool containing the Exchange data begins running before the >clients are done. Sometimes the 'backup stgpool' get stuck waiting >for a tape mount, even though the tape is not mounted, and there are >plenty of tape drives available to be mounted. I scratched my head >about this for awhile, but finally noticed that as soon as the >Exchange client finished it's backup, the tape that was waiting to >mount indeed mounted and the 'backup stgpool' proceeded from there. >Before you say "well, duh!", let me point out that the tape waiting >to mount was not in use by the Exchange client, it had a completely >different tape mounted. >However, the waiting tape was one previously mounted by that >Exchange client during it's backup. > Does the client backup session keep all it's previous tapes it >has used reserved until the session completes? Has anyone seen this >behavior before?
We don't run Exchange, but we have seen similar behavior on sequential access disk volumes when large files were spanned over multiple volumes. Backup storage pool and migration processes could not access the volume containing the first segment of the file until the concurrently running backup session closed the volume containingthe last seqment of the file.
