https://community.hpe.com/t5/operating-system-hp-ux/performance-problem/m-p/2948934#M115533 Most likey causes:<BR /><BR />Too many application/servers running for available membory, swapping is rampant<BR />Too many batch jobs running huge I/O<BR />Incorrect kernel settings.<BR />Short term io hit. <BR />Unbalance disk distribution of a database.<BR /><BR />I assume the situation is more or less constant otherwise you would not be posting.<BR /><BR />I would recommend collecting some data on your system over a longer period of time, and see if the results are conisting. I am attaching just such a script.<BR /><BR />SEP Fri, 11 Apr 2003 06:55:55 GMT Steven E. Protter 2003-04-11T06:55:55Z https://community.hpe.com/t5/operating-system-hp-ux/performance-problem/m-p/2948933#M115532 hi ,experts <BR /> a question gets me . i use 'sar -u 3 5 ' , i found the %wio is especially high .what's the problem ? does %wio impact the performance of the node ? how to improve it ?<BR /> thanks .<BR />------------------------<BR />sar -u 3 5 <BR /><BR />HP-UX pa1022 B.11.00 U 9000/800 04/11/03<BR /><BR />14:58:21 %usr %sys %wio %idle<BR />14:58:24 9 3 77 11<BR />14:58:27 10 2 73 14<BR />14:58:30 7 3 77 14<BR />14:58:33 10 2 71 17<BR />14:58:36 8 2 78 12<BR /><BR />Average 9 2 75 14<BR />---------------------------------<BR />pa1022:/ &gt;vmstat 3 5<BR /> procs memory page faults cpu<BR /> r b w avm free re at pi po fr de sr in sy cs us sy id<BR /> 2 12 0 136319 434604 30 7 0 0 21 0 0 3140 18863 1934 27 4 70<BR /> 1 11 0 136126 434553 7 0 0 0 0 0 0 2971 15928 1228 13 2 85<BR /> 5 11 0 146424 432099 272 20 0 0 1 0 0 2781 20365 1887 54 5 41<BR /> 5 11 0 146424 432872 195 47 0 0 95 0 0 3073 20641 1822 17 2 81<BR /> 0 12 0 142777 433141 173 52 0 0 98 0 0 3271 20712 1690 13 4 83<BR /><BR /> Fri, 11 Apr 2003 06:09:36 GMT https://community.hpe.com/t5/operating-system-hp-ux/performance-problem/m-p/2948933#M115532 miaoxucheng 2003-04-11T06:09:36Z https://community.hpe.com/t5/operating-system-hp-ux/performance-problem/m-p/2948934#M115533 Most likey causes:<BR /><BR />Too many application/servers running for available membory, swapping is rampant<BR />Too many batch jobs running huge I/O<BR />Incorrect kernel settings.<BR />Short term io hit. <BR />Unbalance disk distribution of a database.<BR /><BR />I assume the situation is more or less constant otherwise you would not be posting.<BR /><BR />I would recommend collecting some data on your system over a longer period of time, and see if the results are conisting. I am attaching just such a script.<BR /><BR />SEP Fri, 11 Apr 2003 06:55:55 GMT https://community.hpe.com/t5/operating-system-hp-ux/performance-problem/m-p/2948934#M115533 Steven E. Protter 2003-04-11T06:55:55Z