|
Home > Archive > MS SQL Server > July 2005 > lock timeouts
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
|
|
|
| In our SQL Server, the Profiler is detecting the occaisional Lock:Timeout
event. I am assuming this means that some transaction is blocked, however
the @@LOCK_TIMEOUT value is set to -1 which means that locks never timeout.
So why am I detecting this event?
Any help much appreciated...
| |
| Andrew J. Kelly 2005-07-22, 3:23 am |
| This also includes internal lightweight locks that timeout and are not the
same as a real lock timeout. It's normal and not to be concerned with.
--
Andrew J. Kelly SQL MVP
"Mark" <swozz_@hotmail.com> wrote in message
news:uzFOVYhjFHA.3960@TK2MSFTNGP12.phx.gbl...
> In our SQL Server, the Profiler is detecting the occaisional Lock:Timeout
> event. I am assuming this means that some transaction is blocked, however
> the @@LOCK_TIMEOUT value is set to -1 which means that locks never
> timeout. So why am I detecting this event?
>
> Any help much appreciated...
>
| |
|
| Are you talking about spinlocks here? Isn't it the case that spinlocks never
timeout? Anyway, during these timeouts, SQL Profiler does report an objectID
of 0 most of the time, but occaisionally the object ID is on one of the user
tables - do you still think we shouldn't be concerned with it?
--
Best regards
Mark Baldwin
"Andrew J. Kelly" < sqlmvpnooospam@shadh
awk.com> wrote in message
news:%23QGbifmjFHA.2852@TK2MSFTNGP15.phx.gbl...
> This also includes internal lightweight locks that timeout and are not the
> same as a real lock timeout. It's normal and not to be concerned with.
>
> --
> Andrew J. Kelly SQL MVP
>
>
> "Mark" <swozz_@hotmail.com> wrote in message
> news:uzFOVYhjFHA.3960@TK2MSFTNGP12.phx.gbl...
>
>
| |
| Andrew J. Kelly 2005-07-22, 7:23 am |
| It is hard to say how much of this is caused by the internal stuff or not.
Most likely it is nothing to worry about. But why are you looking at it in
the first place? Did you have an issue specific to locks or timeouts or
just curious? Unless you have an actual problem or other symptoms related
to this I wouldn't worry about it.
--
Andrew J. Kelly SQL MVP
"Mark" <swozz_@hotmail.com> wrote in message
news:%234DOsVrjFHA.1044@tk2msftngp13.phx.gbl...
> Are you talking about spinlocks here? Isn't it the case that spinlocks
> never timeout? Anyway, during these timeouts, SQL Profiler does report an
> objectID of 0 most of the time, but occaisionally the object ID is on one
> of the user tables - do you still think we shouldn't be concerned with it?
>
> --
> Best regards
> Mark Baldwin
>
> "Andrew J. Kelly" < sqlmvpnooospam@shadh
awk.com> wrote in message
> news:%23QGbifmjFHA.2852@TK2MSFTNGP15.phx.gbl...
>
>
| |
|
| We did have a deadlock issue which is now resolved but in the process in
investigating I stumbled on these timeouts messages. Yesterday, the web site
developer says he had one timeout issue, the profiler reported 7 timeout
issues on user data and around 40 timeout issues all together. I am just
trying to understand this Lock:Timeout message to determine if I can use it
to resolve and possibly find as yet unknown problems residing in the
database.
--
Best regards
Mark Baldwin
"Andrew J. Kelly" < sqlmvpnooospam@shadh
awk.com> wrote in message
news:%23xWiDurjFHA.1444@TK2MSFTNGP10.phx.gbl...
> It is hard to say how much of this is caused by the internal stuff or not.
> Most likely it is nothing to worry about. But why are you looking at it
> in the first place? Did you have an issue specific to locks or timeouts
> or just curious? Unless you have an actual problem or other symptoms
> related to this I wouldn't worry about it.
>
> --
> Andrew J. Kelly SQL MVP
>
>
> "Mark" <swozz_@hotmail.com> wrote in message
> news:%234DOsVrjFHA.1044@tk2msftngp13.phx.gbl...
>
>
| |
| David Gugick 2005-07-22, 1:23 pm |
| Mark wrote:
> We did have a deadlock issue which is now resolved but in the process
> in investigating I stumbled on these timeouts messages. Yesterday,
> the web site developer says he had one timeout issue, the profiler
> reported 7 timeout issues on user data and around 40 timeout issues
> all together. I am just trying to understand this Lock:Timeout
> message to determine if I can use it to resolve and possibly find as
> yet unknown problems residing in the database.
>
You can. But as Andrew stated, many of the Timeout errors are internal
lightweight timeouts. You may be able to use the Duration column to
determine if it is a relevant timeout.
--
David Gugick
Quest Software
www.imceda.com
www.quest.com
|
|
|
|
|