That column is text.
Why would you use TEXT in SQL Server 2005? If it is valid XML, why not use
XML? Otherwise, why not use VARCHAR(MAX)? Anyway PageIOLatch_SH is usually
indicative of either poorly performing I/O subsystem (due to hardware) or
pressure from other resources (e.g. memory) that lead to it. Have you
looked at any performance monitor counters other than disk queue length?
What is the difference in configuration between this server and the ones
where this issue doesn't occur? Does this server have the same amount of
physical memory, or less? How is max memory configured? Are both systems
set up for AWE/PAE? What is the amount of RAM installed, and how much is
SQL Server allowed to use? Are they using the same type of disk, or is the
problem system using SAN, NAS, local disk...? Lots and lots of questions,
but without treating them you won't get too many answers, sorry.
On 9/28/08 7:49 PM, in article
2DA391EC-010E-4AB5-B8D0-C70B87D062E4@microsoft.com, "George Shui"
<GeorgeShui@discussions.microsoft.com> wrote:
> That column is text.
>
[quoted text clipped - 3 lines]
>>
>> John
TheSQLGuru - 30 Sep 2008 19:09 GMT
As usual, Aaron is spot-on. RAM and IO.

Signature
Kevin G. Boles
Indicium Resources, Inc.
SQL Server MVP
kgboles a earthlink dt net
> Why would you use TEXT in SQL Server 2005? If it is valid XML, why not
> use
[quoted text clipped - 24 lines]
>>>
>>> John
> That column is text.
>
[quoted text clipped - 3 lines]
>>
>> John
Hi
So there is not really any need for a derived table?
Disk Queue Length is a numeric value so 100% is not a valid value. You don't
say if it a sustained peak or just a peak. Disk Queue length is not always a
reliable value depending on your disc subsystem, but it does seem you do
have a I/O problem see
http://www.microsoft.com/technet/prodtechnol/sql/2005/tsprfprb.mspx. The
first thing I would do is look at what discs you have and what is on them.
John