Log Shipping Problem

Hello Everyone. I am log shippping on a server that has heavy reads and
writes. The Log Ship process runs every 15 minutes and when it runs, it
tends to significantly slow down all the queries running on the box. Queries
which run 20-100 milliseconds run for over 2-3 seconds. The slowing causes
are webservers to back up which in turn slows are website.
I notice in PerMon that after every transaction log backup a checkpoint is
run immedialty after.
The queries are slow during the checkpoint process.
Is there a reason for this? Also, how can I avoid this slow down.
About the Server:
Windows 2003 Enterprise Edition
Quad processeor with Intel XEON 3.0 Ghz.
SQL 2000 version 8.00.936
Thanks,
GeorgeGeorge
There is no real way to avoid the checkpoint process. The checkpoint process
tells SQL Server to write completed transactions to disk, something you
obviously want to happen. You can make the gap between checkpoints longer by
altering the recovery interval. However if you do that the checkpoints will
last longer when they do happen (they will have more to do) and your
transation logs will probably get bigger, and you system will take longer to
come up following a reboot.
About the only thing you can do is look at your disk configuration, to see
if there is a way to speed up your I/O processes.
Hope this helps
John
"George teVelde" wrote:
> Hello Everyone. I am log shippping on a server that has heavy reads and
> writes. The Log Ship process runs every 15 minutes and when it runs, it
> tends to significantly slow down all the queries running on the box. Queries
> which run 20-100 milliseconds run for over 2-3 seconds. The slowing causes
> are webservers to back up which in turn slows are website.
> I notice in PerMon that after every transaction log backup a checkpoint is
> run immedialty after.
> The queries are slow during the checkpoint process.
> Is there a reason for this? Also, how can I avoid this slow down.
> About the Server:
> Windows 2003 Enterprise Edition
> Quad processeor with Intel XEON 3.0 Ghz.
> SQL 2000 version 8.00.936
> Thanks,
> George|||Thanks for the reply, I have one more question.
Is there any reason Log Shipping would grab all the availble threads and not
release them?
I realize that it may not be the log shipping doing it, but I am trying to
eliminate as many variables as I can.
Thanks.|||Hi
No, no reason. If checkpointing affects your performance, the disk layout is
probably not optimal. Basic rule, Logs on a separate set of drive, not on
RAID-5, but on mirrors.
Regards
--
Mike Epprecht, Microsoft SQL Server MVP
Zurich, Switzerland
IM: mike@.epprecht.net
MVP Program: http://www.microsoft.com/mvp
Blog: http://www.msmvps.com/epprecht/
"George teVelde" <GeorgeteVelde@.discussions.microsoft.com> wrote in message
news:69243841-B076-4990-84E1-4C38A7FA4548@.microsoft.com...
> Thanks for the reply, I have one more question.
> Is there any reason Log Shipping would grab all the availble threads and
not
> release them?
> I realize that it may not be the log shipping doing it, but I am trying to
> eliminate as many variables as I can.
> Thanks.