Thursday, February 9, 2012

@@IDLE Bug?

I have written a custom monitor, similar to sp_monitor that stores the
values of @.@.cpu_busy, @.@.io_busy, @.@.idle, etc... to a table.
I have observed that the value of @.@.idle prematurely wraps to a negative
number. Since this value is comprised of a 4 byte integer, one would expect
the value to wrap around to a negative value when it reaches 2,147,483,647.
However, my observations show that this value is wrapping at a value
somewhere just over 441150174 to a negative value somewhere just
under -95648727.
Does anyone have any insight as to why this system variable is prematurely
wrapping? What are the exact values that @.@.idle wraps at?Does anyone have insight into the SQL internals on how the @.@.idle value
rolls over?
"Don Ferguson" <don@.nospamplease> wrote in message
news:uyi0lQLhEHA.1724@.tk2msftngp13.phx.gbl...
> I have written a custom monitor, similar to sp_monitor that stores the
> values of @.@.cpu_busy, @.@.io_busy, @.@.idle, etc... to a table.
> I have observed that the value of @.@.idle prematurely wraps to a negative
> number. Since this value is comprised of a 4 byte integer, one would
expect
> the value to wrap around to a negative value when it reaches
2,147,483,647.
> However, my observations show that this value is wrapping at a value
> somewhere just over 441150174 to a negative value somewhere just
> under -95648727.
> Does anyone have any insight as to why this system variable is prematurely
> wrapping? What are the exact values that @.@.idle wraps at?
>

No comments:

Post a Comment