Values of isreads, iswrites,... in IDS 7.31

Hi all: We used to control the sessions which were runnig with the table "syssesprof" and the fields "isreads", "iswrites" and "isrewrites" (...) when I had IDS 7.30. Now, we have a 7.31 but the values of that fields take values over 1,000,000,000 very quickly. I think these values are not in relation with the querys they are running. Any idea about it? Thanks in advance. Pablo F Herrero Fernandez Gijon Asturias (Spain) mailto:PabloF.Herrero@srrpa.es Sent via Deja.com http://www.deja.com/ Before you buy.

pablof.herrero@srrpa.es wrote: > > Hi all: > We used to control the sessions which were runnig with the table > "syssesprof" and the fields "isreads", "iswrites" and "isrewrites" > (...) when I had IDS 7.30. Now, we have a 7.31 but the values of that > fields take values over 1,000,000,000 very quickly. I think these > values are not in relation with the querys they are running. > Any idea about it? Or many of our queries need to be rewritten, indexes added, and statistics updated. These fields should not behave any differently than they did under 7.31. Are you sure you have not done something careless like drop a zero from your BUFFERS ONCONFIG parameter value so that you are thrashing the buffer pool? What do your cache percentages look like? Art S. Kagel

Hello Art: For example, now my onstat -p tells me: Profile dskreads pagreads bufreads %cached dskwrits pagwrits bufwrits %cached 85355 10381006 541629 84.24 951 1228 14090 93.25 isamtot open start read write rewrite delete commit rollbk 215653 6732 10213 131640 1628 5851 29 79 0 gp_read gp_write gp_rewrt gp_del gp_alloc gp_free gp_curs 0 0 0 0 0 0 0 ovlock ovuserthread ovbuff usercpu syscpu numckpts flushes 0 0 0 108.38 66.48 15 294 bufwaits lokwaits lockreqs deadlks dltouts ckpwaits compress seqscans 5584 0 1344663 0 0 4 6 231 ixda-RA idx-RA da-RA RA-pgsused lchwaits 4294 1 44370 48653 160 And a values's session: isreads iswrites isrewrites 512950274 1947860992 127598592 My BUFFERS ONCONFIG parameter value didn't change from IDS 7.30 and the value is 30000 . In the onstat -p we can read 231 seqscans but the seqscans session value is -263454720. I believe it is a value out of range. I don`t understand anything. Thank you for your interest. Pablo F. Herrero In article <38595A31.E6CBD300@bloomberg.net>, kagel@bloomberg.net wrote: > pablof.herrero@srrpa.es wrote: > > > > Hi all: > > We used to control the sessions which were runnig with the table > > "syssesprof" and the fields "isreads", "iswrites" and "isrewrites" > > (...) when I had IDS 7.30. Now, we have a 7.31 but the values of that > > fields take values over 1,000,000,000 very quickly. I think these > > values are not in relation with the querys they are running. > > Any idea about it? > > Or many of our queries need to be rewritten, indexes added, and statistics > updated. These fields should not behave any differently than they did under > 7.31. Are you sure you have not done something careless like drop a zero > from your BUFFERS ONCONFIG parameter value so that you are thrashing the > buffer pool? What do your cache percentages look like? > > Art S. Kagel > -- Pablo F Herrero Fernandez Gijon Asturias (Spain) Sent via Deja.com http://www.deja.com/ Before you buy.

pablof.herrero@srrpa.es wrote: > "syssesprof" and the fields "isreads", "iswrites" and "isrewrites" > (...) when I had IDS 7.30. Now, we have a 7.31 but the values of that > fields take values over 1,000,000,000 very quickly. I think these > values are not in relation with the querys they are running. > Any idea about it? You must rebuild 'sysmaster' database using 'buildsmi' command i think. > Pablo F Herrero Fernandez Leonid Vorontsov