Topic
No replies
SystemAdmin
SystemAdmin
228 Posts
ACCEPTED ANSWER

Pinned topic IDS 11.50.TC3: ridiculously high time values in wait statistics

‏2008-12-10T12:31:01Z |

SELECT F.TXT[1,10] reason, SUM ( W.WNUM ) numwaits, ( SUM ( W.WCUMTIME )/1000000 ) 
"wait time (s)", SUM ( W.WCUMTIME )/ ( SUM ( W.WNUM ) *1000 ) 
"avg. wait (ms)" FROM SYSMASTER:SYSSCBLST S, SYSMASTER:SYSRSTCB R, SYSMASTER:SYSTWAITS W, SYSMASTER:FLAGS_TEXT F WHERE S.SID = R.SID AND R.TID = W.TID AND W.WREASON = F.FLAGS AND F.TABNAME = 
'systwaits' AND F.TXT  
'condition' AND W.WCUMTIME > 0 and W.WNUM  0 GROUP BY 1 ORDER BY 3 DESC;   reason             numwaits  wait time (s) avg. wait (ms)   running             2801232  1.27027203e14 45346905593.61 aio                  236986 50973701027229 215091613121.6 buffer                47601 15134841176068 317952168569.3 mt yield 0             1873 481058521470.0 256838505856.9 mt yield n              788 360794071221.9 457860496474.4 log i/o                 186 175771382855.1 945007434704.7 mt yield                 99 18502609190.20 186895042325.3 mutex                    54 9251125414.003 171317137296.4 lock                      1 0.283060421264 283.0604212638 checkpoint                1 0.042406745236 42.40674523571   10 row(s) retrieved.


In less than 50 hours (DB up time) my two mostly idle CPUs collect more that 4 million years of "running" time.

Regards
Christian