David,
The underlying backup code is identical for the command-line
and GUI for Data Protection for Exchange.
I agree something doesn't sound right. I would take a closer
look at the logs for both Data Protection for Exchange and
the TSM Server to make sure the same data and same amount
of data was processed in both cases.
Do you know if you used different values for the BUFFERS
and/or BUFFERSIZE? (the DP for Exchange log should show
the values that were used.)
I would try this test again... and see if you see the same thing.
In both backup tests, during the backup, pay attention to the load on
1.) TSM Server 2.) Exchange database disks 3.) Exchange servers 4.)
Network.
If you consistently see this, please call support.
Thanks,
Del
----------------------------------------------------
> We are running Exchange 5.5 on a box running Win2k Sp2, 1.1
GHz
> x2 with 2.6 gig ram, TSM version 5.1.5.2, and TDP version 5.1.5.0. It's
> talking a TSM server running 5.1.6.
>
> The backups are normally ran via a batch file using the
command
> line program with the switches we want, " tdpexcc backup * full
> /tsmoptfile=dsm.opt /logfile=excsch.log >> excfull.log". All paths are
> correct.
>
> The log is telling me that it runs for about 10 hours each
night
> and will send a total of 35 gig. We looked at various things, network,
CPU
> usage, etc. The scheduled backup is ran after 5 PM when most of the
users
> are gone for the day.
>
> We have watched the CPU usage during the day and with just normal
activity
> the usage was 5-10 %. After the backup was started it jumped to 10-20 %
with
> occasional bumps to 80 %. Still not breaking too much of a sweat though.
>
> Then just for the heck of it we ran the same backup via the
GUI
> (tdpexc.exe) and it finished in less than an hour. Same file size, same
> everything else.
>
> We watched the incoming bandwidth on the backup server and
when
> we ran the backup via the GUI it was almost exactly 10 times faster than
> when we ran it via the command line.
>
> In both cases, the full backups were ran during the day with
all
> users online and the backup server itself doing nothing else.
>
> Is something going on here that we are missing?
|
