My company is using a lot of ad hoc queries and I am trying to convince
them to move to stored procedures. However certain managers don't
believe that the compilation of ad hoc queries and the fact that they
usually aren't reused is really anything to worry about. They think
that the cost of doing this is minimal and instead want to focus on
tuning bad queries.
I of course agree that badly written queries should be tuned but it's
also a basic tenet of databases users that stored procedures should
always be used where possible to take advantage of precompilation and
plan caching. However I don't know how to actually prove that with
data.
How can I measure what the overhead is for compiling and building an
execution plan for an ad hoc query vs. the putting the query in a
stored procedure? I've looked at the graphical execution plan for an ad
hoc query and a stored procedure and they look the same.
ThanksSET STATISTICS TIME ON
The first set of data before the query is executed will be the time for
parse and compile. The stats after the data is returned will be the time to
execute.
If the first set of data shows 0, it means the plan is being reused, and
note that SQL Server 2000 can reuse plans for queries other than stored
procedures.
HTH
--
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegroups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>|||In addition to that you need to consider how much memory all those adhoc
plans are taking up in the procedure cache. I have no idea what yours is
like but I have seen systems with all adhoc queries use well over 1GB for
procedure cache on 32 bit systems and larger on 64 bit. That memory can
better be utilized for data and index caching instead of useless plans.
Andrew J. Kelly SQL MVP
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegroups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>|||Thanks for your reply. Would I look at the size of syscacheobjects with
sp_spaceused to determine the amount of memory that the procedure cache
is using?|||Use DBCC MEMORYSTATUS for that:
http://support.microsoft.com/?id=271624
Tibor Karaszi, SQL Server MVP
http://www.karaszi.com/sqlserver/default.asp
http://www.solidqualitylearning.com/
<pshroads@.gmail.com> wrote in message news:1116388073.318788.301230@.f14g2000cwb.googlegroups
.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>|||No, syscacheobjects is a pseudo-table and takes no space on disk, so
sp_spacedused reports 0. The size of the plans is in the pagesused column,
so you can just retrieve the SUM of pagesused.
HTH
--
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116388073.318788.301230@.f14g2000cwb.googlegroups.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>
Showing posts with label execution. Show all posts
Showing posts with label execution. Show all posts
Friday, March 30, 2012
How do I measure the cost of compiling an execution plan?
My company is using a lot of ad hoc queries and I am trying to convince
them to move to stored procedures. However certain managers don't
believe that the compilation of ad hoc queries and the fact that they
usually aren't reused is really anything to worry about. They think
that the cost of doing this is minimal and instead want to focus on
tuning bad queries.
I of course agree that badly written queries should be tuned but it's
also a basic tenet of databases users that stored procedures should
always be used where possible to take advantage of precompilation and
plan caching. However I don't know how to actually prove that with
data.
How can I measure what the overhead is for compiling and building an
execution plan for an ad hoc query vs. the putting the query in a
stored procedure? I've looked at the graphical execution plan for an ad
hoc query and a stored procedure and they look the same.
ThanksSET STATISTICS TIME ON
The first set of data before the query is executed will be the time for
parse and compile. The stats after the data is returned will be the time to
execute.
If the first set of data shows 0, it means the plan is being reused, and
note that SQL Server 2000 can reuse plans for queries other than stored
procedures.
--
HTH
--
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegroups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>|||In addition to that you need to consider how much memory all those adhoc
plans are taking up in the procedure cache. I have no idea what yours is
like but I have seen systems with all adhoc queries use well over 1GB for
procedure cache on 32 bit systems and larger on 64 bit. That memory can
better be utilized for data and index caching instead of useless plans.
--
Andrew J. Kelly SQL MVP
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegroups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>|||Thanks for your reply. Would I look at the size of syscacheobjects with
sp_spaceused to determine the amount of memory that the procedure cache
is using?|||Use DBCC MEMORYSTATUS for that:
http://support.microsoft.com/?id=271624
--
Tibor Karaszi, SQL Server MVP
http://www.karaszi.com/sqlserver/default.asp
http://www.solidqualitylearning.com/
<pshroads@.gmail.com> wrote in message news:1116388073.318788.301230@.f14g2000cwb.googlegroups.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>|||No, syscacheobjects is a pseudo-table and takes no space on disk, so
sp_spacedused reports 0. The size of the plans is in the pagesused column,
so you can just retrieve the SUM of pagesused.
--
HTH
--
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116388073.318788.301230@.f14g2000cwb.googlegroups.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>sql
them to move to stored procedures. However certain managers don't
believe that the compilation of ad hoc queries and the fact that they
usually aren't reused is really anything to worry about. They think
that the cost of doing this is minimal and instead want to focus on
tuning bad queries.
I of course agree that badly written queries should be tuned but it's
also a basic tenet of databases users that stored procedures should
always be used where possible to take advantage of precompilation and
plan caching. However I don't know how to actually prove that with
data.
How can I measure what the overhead is for compiling and building an
execution plan for an ad hoc query vs. the putting the query in a
stored procedure? I've looked at the graphical execution plan for an ad
hoc query and a stored procedure and they look the same.
ThanksSET STATISTICS TIME ON
The first set of data before the query is executed will be the time for
parse and compile. The stats after the data is returned will be the time to
execute.
If the first set of data shows 0, it means the plan is being reused, and
note that SQL Server 2000 can reuse plans for queries other than stored
procedures.
--
HTH
--
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegroups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>|||In addition to that you need to consider how much memory all those adhoc
plans are taking up in the procedure cache. I have no idea what yours is
like but I have seen systems with all adhoc queries use well over 1GB for
procedure cache on 32 bit systems and larger on 64 bit. That memory can
better be utilized for data and index caching instead of useless plans.
--
Andrew J. Kelly SQL MVP
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegroups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>|||Thanks for your reply. Would I look at the size of syscacheobjects with
sp_spaceused to determine the amount of memory that the procedure cache
is using?|||Use DBCC MEMORYSTATUS for that:
http://support.microsoft.com/?id=271624
--
Tibor Karaszi, SQL Server MVP
http://www.karaszi.com/sqlserver/default.asp
http://www.solidqualitylearning.com/
<pshroads@.gmail.com> wrote in message news:1116388073.318788.301230@.f14g2000cwb.googlegroups.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>|||No, syscacheobjects is a pseudo-table and takes no space on disk, so
sp_spacedused reports 0. The size of the plans is in the pagesused column,
so you can just retrieve the SUM of pagesused.
--
HTH
--
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116388073.318788.301230@.f14g2000cwb.googlegroups.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>sql
How do I measure the cost of compiling an execution plan?
My company is using a lot of ad hoc queries and I am trying to convince
them to move to stored procedures. However certain managers don't
believe that the compilation of ad hoc queries and the fact that they
usually aren't reused is really anything to worry about. They think
that the cost of doing this is minimal and instead want to focus on
tuning bad queries.
I of course agree that badly written queries should be tuned but it's
also a basic tenet of databases users that stored procedures should
always be used where possible to take advantage of precompilation and
plan caching. However I don't know how to actually prove that with
data.
How can I measure what the overhead is for compiling and building an
execution plan for an ad hoc query vs. the putting the query in a
stored procedure? I've looked at the graphical execution plan for an ad
hoc query and a stored procedure and they look the same.
Thanks
SET STATISTICS TIME ON
The first set of data before the query is executed will be the time for
parse and compile. The stats after the data is returned will be the time to
execute.
If the first set of data shows 0, it means the plan is being reused, and
note that SQL Server 2000 can reuse plans for queries other than stored
procedures.
HTH
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegr oups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>
|||In addition to that you need to consider how much memory all those adhoc
plans are taking up in the procedure cache. I have no idea what yours is
like but I have seen systems with all adhoc queries use well over 1GB for
procedure cache on 32 bit systems and larger on 64 bit. That memory can
better be utilized for data and index caching instead of useless plans.
Andrew J. Kelly SQL MVP
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegr oups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>
|||Thanks for your reply. Would I look at the size of syscacheobjects with
sp_spaceused to determine the amount of memory that the procedure cache
is using?
|||Use DBCC MEMORYSTATUS for that:
http://support.microsoft.com/?id=271624
Tibor Karaszi, SQL Server MVP
http://www.karaszi.com/sqlserver/default.asp
http://www.solidqualitylearning.com/
<pshroads@.gmail.com> wrote in message news:1116388073.318788.301230@.f14g2000cwb.googlegr oups.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>
|||No, syscacheobjects is a pseudo-table and takes no space on disk, so
sp_spacedused reports 0. The size of the plans is in the pagesused column,
so you can just retrieve the SUM of pagesused.
HTH
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116388073.318788.301230@.f14g2000cwb.googlegr oups.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>
them to move to stored procedures. However certain managers don't
believe that the compilation of ad hoc queries and the fact that they
usually aren't reused is really anything to worry about. They think
that the cost of doing this is minimal and instead want to focus on
tuning bad queries.
I of course agree that badly written queries should be tuned but it's
also a basic tenet of databases users that stored procedures should
always be used where possible to take advantage of precompilation and
plan caching. However I don't know how to actually prove that with
data.
How can I measure what the overhead is for compiling and building an
execution plan for an ad hoc query vs. the putting the query in a
stored procedure? I've looked at the graphical execution plan for an ad
hoc query and a stored procedure and they look the same.
Thanks
SET STATISTICS TIME ON
The first set of data before the query is executed will be the time for
parse and compile. The stats after the data is returned will be the time to
execute.
If the first set of data shows 0, it means the plan is being reused, and
note that SQL Server 2000 can reuse plans for queries other than stored
procedures.
HTH
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegr oups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>
|||In addition to that you need to consider how much memory all those adhoc
plans are taking up in the procedure cache. I have no idea what yours is
like but I have seen systems with all adhoc queries use well over 1GB for
procedure cache on 32 bit systems and larger on 64 bit. That memory can
better be utilized for data and index caching instead of useless plans.
Andrew J. Kelly SQL MVP
<pshroads@.gmail.com> wrote in message
news:1116376117.157139.259730@.g44g2000cwa.googlegr oups.com...
> My company is using a lot of ad hoc queries and I am trying to convince
> them to move to stored procedures. However certain managers don't
> believe that the compilation of ad hoc queries and the fact that they
> usually aren't reused is really anything to worry about. They think
> that the cost of doing this is minimal and instead want to focus on
> tuning bad queries.
> I of course agree that badly written queries should be tuned but it's
> also a basic tenet of databases users that stored procedures should
> always be used where possible to take advantage of precompilation and
> plan caching. However I don't know how to actually prove that with
> data.
> How can I measure what the overhead is for compiling and building an
> execution plan for an ad hoc query vs. the putting the query in a
> stored procedure? I've looked at the graphical execution plan for an ad
> hoc query and a stored procedure and they look the same.
> Thanks
>
|||Thanks for your reply. Would I look at the size of syscacheobjects with
sp_spaceused to determine the amount of memory that the procedure cache
is using?
|||Use DBCC MEMORYSTATUS for that:
http://support.microsoft.com/?id=271624
Tibor Karaszi, SQL Server MVP
http://www.karaszi.com/sqlserver/default.asp
http://www.solidqualitylearning.com/
<pshroads@.gmail.com> wrote in message news:1116388073.318788.301230@.f14g2000cwb.googlegr oups.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>
|||No, syscacheobjects is a pseudo-table and takes no space on disk, so
sp_spacedused reports 0. The size of the plans is in the pagesused column,
so you can just retrieve the SUM of pagesused.
HTH
Kalen Delaney
SQL Server MVP
www.SolidQualityLearning.com
<pshroads@.gmail.com> wrote in message
news:1116388073.318788.301230@.f14g2000cwb.googlegr oups.com...
> Thanks for your reply. Would I look at the size of syscacheobjects with
> sp_spaceused to determine the amount of memory that the procedure cache
> is using?
>
Wednesday, March 7, 2012
How do I figure out where SSIS logs are?
Dear all,
I was wondering how to open .LOG files created for a SSIS execution package.
Thanks in advance,
Open them in a text editor. I'm a bag fan of Textpad (www.textpad.com)
You control where they are. You tell tell the package where to log to.
-Jamie
|||Hi Jamie,
Come on man, textpad? Is my favourite without a doubt, I like it a lot too.
I was saying, in what path ssis leave its logs. where?
Thanks again
|||Like I said previous, the logs are put wherever you tell SSIS to put them. That can be a text file (it is you, not SSIS, that defines where that file resides), SQL Server (in which case they'll be in the msdn..sysdtspackageslog90 table if memory serves), event log, SQL Profiler or an XML file.
-Jamie
Subscribe to:
Posts (Atom)