I haven't seen this on the forums yet so I figured I'd ask.
Weekly we have one large (1tb or more) object, while monthly we have 3 large (again, 1tb or more) objects that have a very odd behavior I've not observed on other backup object.
Backups to our VLS run as normal. Once the jobs complete and we start our copies to tape (automated, scheduled), the three largest objects will begin copying amongs the earliest object. If watched, they will reach anywhere between 65% and 90% before 'resetting' back to anywhere between 45% and 50% and continuing. It usually coincides with loading a new virtual tape to continue reading data. A new tape will load on the VLS, the progress of these large jobs will reduce down. I've personally watched this happen 4-5 times per copy job each time the objects need to be copied.
Any ideas? Am I just seeing the system estimate an objects completion then changing it once it reads more data off a new virtual tape? Or am I seeing an actual error writing data that's causing re-writes? Error logs do not indicate any error conditions.
HPDP version 6.11
VLS: SW 6200 series running 3.4.1.1
Destination devices MSL G3 running 8.60 / drives HP Ultrium 3 running L63W