Re: Tunning question

In article <6so5p8$rs5$1@nnrp1.dejanews.com>, PabloF@my-dejanews.com () wrote: > > > Hi all, > we have a 2 CPU, Unix Alpha 8200 machine running Informix ODS 7.23.FC4. > onstat -p shows:> dskreads pagreads bufreads %cached dskwrits pagwrits bufwrits %cached > 4966206 17506603 190285614 97.39 338155 613077 18712111 98.19 > > isamtot open start read write rewrite delete commit > rollbk > 105442498 3934478 7887443 58928336 4183669 36989 118823 5868 > 6 > > ovlock ovuserthread ovbuff usercpu syscpu numckpts flushes > 0 0 0 16635.87 9806.50 377 1772 > > bufwaits lokwaits lockreqs deadlks dltouts ckpwaits compress seqscans > 726885 0 102621654 0 0 309 42825 362111 > > ixda-RA idx-RA da-RA RA-pgsused lchwaits > 1739641 204790 1316146 3251917 45756 > > onstat -g ioq shows:> > AIO I/O queues: > q name/id len maxlen totalops dskread dskwrite dskcopy > kio 0 0 0 0 0 0 0 > kio 1 0 16 5304499 4966351 338148 0 > adt 0 0 0 0 0 0 0 > msc 0 0 2 27466 0 0 0 > aio 0 0 2 16282 10346 0 0 > pio 0 0 0 0 0 0 0 > lio 0 0 1 1 0 1 0 > gfd 3 0 6 40 37 3 0 > gfd 4 0 6 40 37 3 0 > gfd 5 0 3 4 1 3 0 > gfd 6 0 1 3 3 0 0 > gfd 7 0 1 1 1 0 0 > gfd 8 0 1 1 1 0 0 > gfd 9 0 1 3 3 0 0 > gfd 10 0 1 1 1 0 0 > gfd 11 0 1 3 3 0 0 > gfd 12 0 1 1 1 0 0 > gfd 13 0 1 1 1 0 0 > > and finally onstat -g iof shows: > AIO global files: > gfd pathname totalops dskread dskwrite io/s > 3 /dev/chunk_reca1 6472 3604 2868 0.0 > 4 /dev/chunk_reca1 1404 21 1383 0.0 > 5 /dev/chunk_reca2 152759 58016 94743 0.6 > 6 chunk_reca31 185274 176809 8465 0.7 > 7 chunk_reca32 393653 367006 26647 1.5 > 8 chunk_reca33 215259 194982 20277 0.8 > 9 chunk_reca41 179143 175659 3484 0.7 > 10 chunk_reca42 103818 103786 32 0.4 > 11 chunk_reca51 1485634 1482189 3445 5.6 > 12 chunk_reca52 311650 293167 18483 1.2 > 13 chunk_reca53 611404 610311 1093 2.3 > > Every day, there are peaks where the alpha is very slowly. > We have the SHARED MEMORY PARAMETERS > Server Number [ 0] Server Name > [onlinereca_soc ] > Server Aliases [ > ] > Dbspace Temp [tempdbs > ] > Deadlock Timeout [ 60] Secs Number of Page Cleaners > [ 4] > Forced Residency [N] Stack Size (K) > [ 64] > Non Res. SegSize (K) [ 90000] Optical Cache Size (K) [ > 0] > > Dbspace Down Option > [0] > Preserve Log For Log > Backup [N] > Heterogeneous Commit [N] Transaction Timeout > [ 300] > Physical Log Buffer Size [ 256] K Long TX HWM > [ 45] > Logical Log Buffer Size [ 32] K Long TX HWM Exclusive > [ 50] > Max # of Logical Logs [ 20] Index Page Fill Factor > [ 90] > Max # of Locks [ 50000] Add SegSize (K) [ > 20000] > Max # of Buffers [ 40000] Total Memory (K) [ > 0] > > Resident Shared Memory size [ 100348] Kbytes Page Size [ 2] > Kbytes > [cutting] chunk_reca51 seems to be taking a lot of hits. I'd increase the number of tempdbs as well, but I like lots of tempdbs across lots of disks and controllers Paul Watson # I don't suffer from WF Software Ltd. # stress, I'm just Tel. (+44) 1436 674729 # a carrier Fax. (+44) 1436 678693 # www.wfsoftware.co.uk