|
|
Question : Waittype PAGEIOLATCH_SH with a large waittime is degrading system performance
|
|
This morning system performance was degrading as the number of the system processes was exploding. Normally I have about 20 processes running on a database (master..sysprocesses), but in one hour it grew to 200. Wheb I checked the table I saw this:
spid wait waittime lastwaittype trans status 54 0x0422 262937 PAGEIOLATCH_SH 0 sleeping (about 140 of these) 61 0x0422 28304 PAGEIOLATCH_SH 2 sleeping (about 20 of these) 66 0x0000 0 PAGELATCH_UP 2 runnable (2 of these) 88 0x0000 0 PAGEIOLATCH_EX 2 runnable (1 of these) 101 0x0424 28187 PAGEIOLATCH_EX 2 sleeping (4 of these)
I restarted the server and the problem was solved. All the processes came from the .Net SqlClient Data Provider. Some of my SQL jobs got the following warning during that hour:
Executed as user: NETMATCH5\sqlserver. Time-out occurred while waiting for buffer latch type 2 for page (1:147380), database ID 10. [SQLSTATE 42000] (Error 845). The step failed.
My questions are: 1. What do the waittype 0x0422 and 0x0424 mean? I looked on internet but could not find a list of waittypes. 2. What could the reason be, disk IO too much stress?
Regards, Lex van de Pol
|
Answer : Waittype PAGEIOLATCH_SH with a large waittime is degrading system performance
|
|
I had similar errors. While pondering their significance for a day, the RAID controller card died. I don't know if this was coincidence or the errors were trying to warn me badness was about to happen.
|
|
|
|
|