Hello All,
I'm trying to explain why my level 0 rman backups have suddenly jumped from
8hours to 12 hours. The size of the backup has barely changed and there has
been no configuration changes. AWR and O/S statistics show that CPU for all
the channels are maxed out at ~99% (presumably backup compression) for most of
this time, both in the 8 hour case and 12 hour case. The server itself has no
other activity during the backup window.
This is on 11.2.0.4 on AWS EC2 backing up to S3 storage.
I'm not looking for tuning advise here (thanks though!) - I'm really hoping for
some advise on finding an explanation for the change in times. Given that all
it's doing is sitting on CPU I'm not sure what else I can investigate -
especially with limited ability to cross-reference it with the "good old" runs.
It's been 8 hours for months and months, and has now been 12 hours for the
last 3 weeks. There was no gradual increase - just a step change.
Thanks!Charlotte