> >From your statistics I see that your log is 2000 MB:
> Log:
>Available Space (MB): 2,000
> Assigned Capacity (MB): 2,000
> Maximum Extension (MB): 0
> Maximum Reduction (MB): 1,996
> Page Size (bytes): 4,096
> Total Usable Pages: 511,488
>
I've had a log max out at 5GB in normal - due to one very slow running
client (love those duplex mismatches!). The client backup ran almost all
day, and we were doing some Delete Filespaces, which hit the log/db pretty
hard. Since the log is flushed FIFO (see note), and the client session
hadn'
>I'm curious about one thing though. When in "normal" mode,
>log consumption
>rarely should exceed about 1%. Yet your statistics show a max
>utilization
>of 89.3%. How did you do that in NORMAL mode?
Tab,
This is an excellent question. I was out all last week so I'm not sure what
happened. I
Geoffrey,
Switching to roll-forward mode can be done at any time. As Wanda said, it
kicks off a full database backup to allow it to zero the log at a known
point.
Once in roll-forward mode, database changes will accumlate in the recovery
log until the next DB backup so log consumption will go u
You can reset the cumulative consumption field in the q logvol f=d display.
Then run your server normally for 24 hours and check the consumption. This
will give you a good idea how much recover log you might use in a 24 hour
period (the period between scheduled db backups).
The command to reset
Yes, it does no harm to switch to ROLLFORWARD mode in mid-stream, it just
works.
It does fire an extra DB backup; or maybe that's when you switch BACK from
ROLLFORWARD to NORMAL, I forget.
The amount of log space you need depends on the amount of activity rather
than the DB size, so it's hard to
Hi Geoff,
You can increase BUFPOOLSIZE only if it take REAL memory, not swap.
use the SVMON -P to check how much RAM TSM is really using.
Before turning TSM into roll forward, reset your LOG MAX % UTIL and
wait for a day or two to check the max % used. (adsm> help reset log)
2 Gig seems to be
