This thread has gone around a lot and evolved, so here's my 2 cents on it.
It's sad to be thinking that 44G is too big for a database. I subscribe to
the db2 list and tables that are 100's of Gigs are discussed regularly,
never mind whole databases. I urge the people who attend the
upcoming synposium in the U.K. to really jump on IBM about this!
All sequential DB operations must be multi-threaded, for instance
a full db backup should be able to write to 5 tapes at once.
You will have the right person there, Mike Kaczmarski, to make this
case to.
Richard Simms replied that people should call any & all problems
into IBM support. I can attest that it works. the original
post noted that message number ANR4556W was issued. Why did this
message come out? Because the first time this problem happend to me
I called IBM support. There were 2 issues for me form the start; first that
it happened at all and 2nd that there wa no message to say that the
incremental db backup didn't produce the expected results, that is, 0% log
usage. I eventaully understood that the problem can happen because of
long running transactions, and that it can't easily be completely eliminated,
but I insisted that it was a bug that there was no message.
Now I use the message to respond to the situation. I run the trigger at 25%.
If the message comes out, it is trapped and the trigger is raised to 75% to
stop the loop of incrementals. A page is also sent to my beeper so I can
find and fix the root cause of the problem.
(The server is on os/390, level 3.1.2.40.
I use OPS/MVS to trap the message. The used db pages
is 9,428,172 and growing. The log has 599,552 pages total.
I have 1 server and no plans to split it or start a
2nd server).
--
--------------------------
--------------------------
Bill Colwell
Bill Colwell
C. S. Draper Lab
Cambridge, Ma.
bcolwell AT draper DOT com
--------------------------
In <Pine.GSO.4.05.9909022250090.18685-100000 AT rexx.acsu.buffalo DOT edu>, on
09/02/99
In <Pine.GSO.4.05.9909022250090.18685-100000 AT rexx.acsu.buffalo DOT edu>, on
09/02/99
at 10:58 PM, Steven P Roder <tkssteve AT REXX.ACSU.BUFFALO DOT EDU> said:
>>
>> How big is your database? How much does it change in a day? I've never
>> seen a log that big fill without operational problems preventing the
>> regular backup or the triggered backup to fail...
>My DB is too big, and that is why I am in the process of setting up a
>second server, this time under AIX (primary one is Sun with 2 3575-L12's
>with 4 3570-C1A's each) with a 3494 and 4 3590-B1A drives.
>q db
>Available Assigned Maximum Maximum Page Total Used Pct
>Max.
> Space Capacity Extension Reduction Size Usable Pages Util
>Pct
> (MB) (MB) (MB) (MB) (bytes) Pages
>Util
>--------- -------- --------- --------- ------- --------- --------- -----
>-----
> 67,780 56,000 11,780 12,436 4,096 14,336,00 11,147,61 77.8
>77.8
> 0 3
>On 8/22, it had 10,868,322
> 8/29, it was 11,031,226
>..but, school is back in session. The biggest problem is running tasks
>like Expiration or reclaimation (I do move data's "by Hand"). Expiration
>takes about 4 days, so I can only make one pass/week. The restartable
>expiration in 3.1.2.30/40 is a lifesaver for me.
>Full backup is done to DLT4000 and takes about 3:45 on 8/29.
>Steve (unVMix Systems Programmer/Dude) Roder
>(tkssteve AT ubvm.cc.buffalo DOT edu | tkssteve AT acsu.buffalo DOT edu |
>(716)645-3564 ,
> | http://ubvm.cc.buffalo.edu/~tkssteve)
|