n/w issues and had to reboot.

Has anyone sees issues where although the exec plans are the same , theres
no blocking, and identical hardware that the query would perform slower at
times to return the results to the client whereas it runs much faster when
run on the same box. The query return a few million rows.. It takes as long
as 1 hr when its slow and as fast as 5 mins when its normal to obtain these
results on the client.
Even stopping and starting SQL doesnt work. I just have to reboot and then
its all fast again. I am thinking its being choked someplace.
I have seen this same behaviour on another server as well. and only a reboot
would work. No messages in the log, nothing..
A reboot always solves it. Any idea why.. Using SQL 2000 on Windows 2003
Try sp_updatestats
I had simillar problem, VB6 client ran very slow sometimes, but the same
query (meanwhile!) ran fast in QA.
tv
Hassan napsal(a):
> Has anyone sees issues where although the exec plans are the same , theres
> no blocking, and identical hardware that the query would perform slower at
> times to return the results to the client whereas it runs much faster when
> run on the same box. The query return a few million rows.. It takes as long
> as 1 hr when its slow and as fast as 5 mins when its normal to obtain these
> results on the client.
> Even stopping and starting SQL doesnt work. I just have to reboot and then
> its all fast again. I am thinking its being choked someplace.
> I have seen this same behaviour on another server as well. and only a reboot
> would work. No messages in the log, nothing..
> A reboot always solves it. Any idea why.. Using SQL 2000 on Windows 2003
>
>
|||Have your done profiling and perf-mon'ing of the boxes during these phases?
What are your cache hit ratio's like? Are you suffereing from disk IO queue's
going excessive? Are you having excessive paging? What does CPU utilization
look like?
When it takes 5 minutes, is it after a few iterations (i.e. data has been
cached by previous runs), etc.
"Hassan" wrote:

> Has anyone sees issues where although the exec plans are the same , theres
> no blocking, and identical hardware that the query would perform slower at
> times to return the results to the client whereas it runs much faster when
> run on the same box. The query return a few million rows.. It takes as long
> as 1 hr when its slow and as fast as 5 mins when its normal to obtain these
> results on the client.
> Even stopping and starting SQL doesnt work. I just have to reboot and then
> its all fast again. I am thinking its being choked someplace.
> I have seen this same behaviour on another server as well. and only a reboot
> would work. No messages in the log, nothing..
> A reboot always solves it. Any idea why.. Using SQL 2000 on Windows 2003
>
>
|||Even a restart of SQL Service wont help.. But a reboot of a server sure
does fix it
"Wanderer" <Wanderer@.discussions.microsoft.com> wrote in message
news:957C21BB-A9EE-4E33-8903-205053E55821@.microsoft.com...
> Have your done profiling and perf-mon'ing of the boxes during these
phases?
> What are your cache hit ratio's like? Are you suffereing from disk IO
queue's
> going excessive? Are you having excessive paging? What does CPU
utilization[vbcol=seagreen]
> look like?
> When it takes 5 minutes, is it after a few iterations (i.e. data has been
> cached by previous runs), etc.
> "Hassan" wrote:
theres[vbcol=seagreen]
at[vbcol=seagreen]
when[vbcol=seagreen]
long[vbcol=seagreen]
these[vbcol=seagreen]
then[vbcol=seagreen]
reboot[vbcol=seagreen]