Message-Id: <10583.114037@fatcity.com> From: Rahul Date: Tue, 8 Aug 2000 14:28:43 +0700 Subject: RE: urgent bstat/estat performance tuning reads and blks_ reads reads = number of read requests sent blk_reads = total number of blocks read blk_read / reads will give you an idea of how many blks read per read. if the blks_read figure is close to reads then sequential scans are happening.. if the "per read" figure is closer to ur multiblock read then FTS are going on. > ---------- > From: Sairlao, Chark[SMTP:SairlaoC@transfield.com.au] > Sent: Tuesday, August 08, 2000 2:19 PM > To: 'oracledba@quickdoc.co.uk' > Subject: urgent *********** bstat/estat performance tuning reads and blks_ reads > > hi gurus, > > what is reads , blks_read, writes, blks_writes meant in the report?. > > And to rebalance the disk what figures should I be looking at read_time, > write_time or megaBytes. > > I know the two disks is pretty balance at the moment, users are complaining > it is slow looking at the report , db file sequential is accounted for 50% > of the wait and db file parallel write accounted for 30% of the > wait,therefore I will be suggesting for 2 more disks to split out the > data01.dbf and idx01.dbf, it should speed up a fair bit is this right? > > I know log file sync is accounted for 10% of the wait, but there is not much > I can do as it is baan application, and developers does not want to the the > code to reduce the commit statements. > > > > 14> order by n1.name; > Statistic Total Per Transact Per Logon Per > Second > --------------------------- ------------ ------------ ------------ > ------------ > CPU used by this session 278685 18 397.55 > 9.68 > CPU used when call started 278675 18 397.54 > 9.68 > CR blocks created 1395 .09 1.99 > .05 > Current blocks converted fo 20 0 .03 > 0 > DBWR buffers scanned 2679892 173.11 3822.96 > 93.06 > > SVRMGR> select n1.event "Event Name", > 2> n1.event_count "Count", > 3> n1.time_waited "Total Time", > 4> round(n1.time_waited/n1.event_count, 2) "Avg Time" > 5> from stats$event n1 > 6> where n1.event_count > 0 > 7> order by n1.time_waited desc; > Event Name Count Total Time Avg Time > -------------------------------- ------------- ------------- ------------- > SQL*Net message from client 2132889 126070305 59.11 > db file sequential read 1104139 167380 .15 > log file sync 15779 30771 1.95 > SQL*Net message to client 2132904 3615 0 > write complete waits 284 3550 12.5 > free buffer waits 738 1527 2.07 > db file scattered read 3051 1431 .47 > latch free 707 993 1.4 > log file switch completion 10 642 64.2 > rdbms ipc reply 135 156 1.16 > > > VRMGR> Rem System wide wait events for background processes (PMON, SMON, > etc) > SVRMGR> select n1.event "Event Name", > 2> n1.event_count "Count", > 3> n1.time_waited "Total Time", > 4> round(n1.time_waited/n1.event_count, 2) "Avg Time" > 5> from stats$bck_event n1 > 6> where n1.event_count > 0 > 7> order by n1.time_waited desc; > Event Name Count Total Time Avg Time > -------------------------------- ------------- ------------- ------------- > rdbms ipc message 47028 11376404 241.91 > smon timer 115 2886213 25097.5 > pmon timer 9568 2880016 301.01 > db file parallel write 4979 84764 17.02> > log file parallel write 16926 33353 1.97 > control file parallel write 360 1680 4.67 > db file scattered read 2652 1417 .53 > db file single write 207 672 3.25 > log file sync 238 584 2.45 > db file sequential read 599 302 .5 > latch free 94 190 2.02 > > > 5> from stats$files order by table_space, file_name; > TABLE_SPACE FILE_NAME > READS BLKS_READ READ_TIME WRITES BLKS_WRT WRITE_TIME MEGABYTES > ------------------------------ > ------------------------------------------------ ---------- ---------- > ---------- ---------- ---------- ---------- ---------- > DATSPACE /u101/oradata/baanIV/data01.dbf > 333534 336221 45529 2295 2295 7407 1573 > DATSPACE /u101/oradata/baanIV/data04.dbf > 66329 66329 6397 94 94 346 157 > DATSPACE /u101/oradata/baanIV/data05.dbf > 91727 91727 9090 871 871 2383 524 > DATSPACE /u101/oradata/baanIV/db02.dbf > 10179 10179 3004 370 370 1213 157 > DATSPACE /u101/oradata/baanIV/db03.dbf > 90122 90122 7853 723 723 2196 157 > DATSPACE /u101/oradata/baanIV/db05.dbf > 294734 294734 32418 860 860 2662 524 > IDXSPACE /u100/oradata/baanIV/idx01.dbf > 70153 70153 22742 5581 5581 16673 1332 > IDXSPACE /u100/oradata/baanIV/idx02.dbf > 15193 15193 3700 478 478 1388 105 > IDXSPACE /u100/oradata/baanIV/idx03.dbf > 7122 7122 3116 602 602 2028 105 > IDXSPACE /u100/oradata/baanIV/idx04.dbf > 14031 14031 2033 108 108 413 105 > IDXSPACE /u100/oradata/baanIV/idx05.dbf > 6427 6427 3161 706 706 2562 157 > IDXSPACE /u100/oradata/baanIV/idx06.dbf > 25142 25142 7252 1857 1857 5707 524 > IDXSPACE /u100/oradata/baanIV/idx07.dbf > 15366 15366 2818 393 393 1252 105 > IDXSPACE /u101/oradata/baanIV/db06.dbf > 53150 53150 16831 1495 1495 4738 524 > > any comments are welcome > > Chark > > > -------- > If you're bored, then visit the list's website: http://www.lazydba.com (updated daily) > to unsubscribe, send a blank email to oracledba-unsubscribe@quickdoc.co.uk