July 17, 2003 at 7:32 am
Hi!
While monitoring MS SQL 2000 sp3 + (Win 2000 Adv. Server sp4) noticed, that Processor time counter of a big 6xXeon 700 system drops in few seconds from ~80% to 0% in the middle of a work time. Manual restart corrected the problem, but I am very curious about what has happened? Just before this the following error happened several times:
(no other error, that could address this problem, neither in windows log nor in ms sql error logs were found)
Event Type:Warning
Event Source:Npfs
Event Category:None
Event ID:54
Date:17.07.2003
Time:13:45:42
User:N/A
Computer:JUPITER
Description:
An Io Request to the device \Device\NamedPipe did not complete or canceled within the specific timeout. This can occur if the device driver does not set a cancel routine for a given IO request packet.
Data:
0000: 00 00 00 00 01 00 52 00 ......R.
0008: 00 00 00 00 36 00 04 80 ....6..€
0010: 00 00 00 00 00 00 00 00 ........
0018: 00 00 00 00 00 00 00 00 ........
0020: 00 00 00 00 00 00 00 00 ........
July 21, 2003 at 8:00 am
This was removed by the editor as SPAM
July 23, 2003 at 9:49 am
Just a thought... we were using named-pipes and had network time-out issues. Named pipes doesn't support buffering the same way TCP/IP does. I would recommend using only TCP/IP if at all possible. Look at the SQL Books Online for "Named Pipes, Named Pipes vs. TCP/IP Sockets"
-Dan
-Dan
July 23, 2003 at 9:59 am
" I would recommend using only TCP/IP if at all possible."
- Not possible at this time, too many apps to rewrite or change.
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply