Running a LIKE statement when searching for a date field...

I am trying to run a like statement that has a datetime column and for some reason it does not return any values. I looked in the SQL help files and in states in there that when trying to select using a datetime that the preferred way of doing it is using a like statment. Does anybody know a better way of doing this? Here is my example: (I have dates in this column ie 2006-02-13 11:30:54.220)

SELECT * FROM workorderhistory WHERE wheninstalled LIKE '%2006-02%'

Where did you find that preferred way to compare dates is LIKE. You might want to use many date functions to compare dates available in 2k and 2k5.

OR if you incist on using LIKE convert "wheninstalled" to format that you specify in LIKE.

|||datetime is stored in database in an internal format. It is not in YYYY-MM-DD or whatsoever format.

>> I have dates in this column ie 2006-02-13 11:30:54.220
This just how Query Analyser represent the date or format the date and time when it return the records.

>> SELECT * FROM workorderhistory WHERE wheninstalled LIKE '%2006-02%'
To retireve records for month of Feb 2006,

SELECT * FROM workorderhistory WHERE wheninstalled >= '2006-02-01' and wheninstalled < '2006-03-01'

|||

Here is the text from the sql help file.

It is recommended that LIKE be used when you search for datetime values, because datetime entries can contain a variety of dateparts. For example, if you insert the value 19981231 9:20 into a column named arrival_time, the clause WHERE arrival_time = 9:20 cannot find an exact match for the 9:20 string because SQL Server converts it to Jan 1, 1900 9:20AM. A match is found, however, by the clause WHERE arrival_time LIKE '%9:20%'.

Thanks for your help.

|||Thanks for the help... I don't konw why I did not think of doing it that way... Slipped my mind I guess.|||

try this may be helpful for you

SELECT * FROM
WHERE (CAST(FLOOR(CAST([date] AS FLOAT)) AS DATETIME) = '3/14/2006')

|||

How can I use a like statement in there. I tried doing

select * from workorderhistory where (CAST(FLOOR(CAST([date] AS FLOAT)) AS DATETIME) like '03/%') and this did not work. Basically I want to see all the workorders that were installed in the month of march.

Running a LIKE statement when searching for a date field...

I am trying to run a like statement that has a datetime column and for some reason it does not return any values. I looked in the SQL help files and in states in there that when trying to select using a datetime that the preferred way of doing it is using a like statment. Does anybody know a better way of doing this? Here is my example: (I have dates in this column ie 2006-02-13 11:30:54.220)

SELECT * FROM workorderhistory WHERE wheninstalled LIKE '%2006-02%'

Where did you find that preferred way to compare dates is LIKE. You might want to use many date functions to compare dates available in 2k and 2k5.

OR if you incist on using LIKE convert "wheninstalled" to format that you specify in LIKE.

|||datetime is stored in database in an internal format. It is not in YYYY-MM-DD or whatsoever format.

>> I have dates in this column ie 2006-02-13 11:30:54.220
This just how Query Analyser represent the date or format the date and time when it return the records.

>> SELECT * FROM workorderhistory WHERE wheninstalled LIKE '%2006-02%'
To retireve records for month of Feb 2006,

SELECT * FROM workorderhistory WHERE wheninstalled >= '2006-02-01' and wheninstalled < '2006-03-01'

|||

Here is the text from the sql help file.

It is recommended that LIKE be used when you search for datetime values, because datetime entries can contain a variety of dateparts. For example, if you insert the value 19981231 9:20 into a column named arrival_time, the clause WHERE arrival_time = 9:20 cannot find an exact match for the 9:20 string because SQL Server converts it to Jan 1, 1900 9:20AM. A match is found, however, by the clause WHERE arrival_time LIKE '%9:20%'.

Thanks for your help.

|||Thanks for the help... I don't konw why I did not think of doing it that way... Slipped my mind I guess.|||

try this may be helpful for you

SELECT * FROM
WHERE (CAST(FLOOR(CAST([date] AS FLOAT)) AS DATETIME) = '3/14/2006')

|||

How can I use a like statement in there. I tried doing

select * from workorderhistory where (CAST(FLOOR(CAST([date] AS FLOAT)) AS DATETIME) like '03/%') and this did not work. Basically I want to see all the workorders that were installed in the month of march.

Runniing Values in Charts

I have a table which show the sales year to date by week. The sales amount is
a running value and it will keep incrementing as the weeks progress.
If i use the same value for the chart it does not allow RunningValue for
scope Nothing. How do i do the equalant in a chart to xhow the cumulative by
week rather than just the week figures. ex:
Actual Cumulative
Week1 1000 1000
Week2 1500 2500
Week3 1000 3500
First column is
Sum(cdec( Fields!CURRENTANNUALPREMIUM.Value ))
second is
RunningValue(cdec( Fields!CURRENTANNUALPREMIUM.Value ), SUM, Nothing)
this works fine on table but not in chart. Please help me do cumulative charts
regards
GaneshSorry, RunningValue() in charts is not supported on RS 2000. It is supported
on RS 2005 though.
-- Robert
This posting is provided "AS IS" with no warranties, and confers no rights.
"Ganesh" <Ganesh@.discussions.microsoft.com> wrote in message
news:25E16A7D-A53D-4DFE-BD5E-4036A3C56EBD@.microsoft.com...
>I have a table which show the sales year to date by week. The sales amount
>is
> a running value and it will keep incrementing as the weeks progress.
> If i use the same value for the chart it does not allow RunningValue for
> scope Nothing. How do i do the equalant in a chart to xhow the cumulative
> by
> week rather than just the week figures. ex:
> Actual Cumulative
> Week1 1000 1000
> Week2 1500 2500
> Week3 1000 3500
> First column is
> Sum(cdec( Fields!CURRENTANNUALPREMIUM.Value ))
> second is
> RunningValue(cdec( Fields!CURRENTANNUALPREMIUM.Value ), SUM, Nothing)
> this works fine on table but not in chart. Please help me do cumulative
> charts
> regards
> Ganesh

Run two stored procs in parallel

Hi
I have a simple stored proc which runs in less than a second.
This stored proc calculates different values based on parameters
passed.
For example sp_name "test 1 ' and
sp_name 'test 2 ' are totally different.
Now The issue is this stored proc is called more than 500 times in some
applications.
So these applications run for 500+ seconds . Although load on sql
server is almost negligible. Memory is also hugely available.
I need your suggestion in running these all at the same time.
In other words all 500 must be started at the same time ..there is no
need for one procedure to complete and the next to start. None of the
results are depenedent on any other results.
pls help me .
KannanIs this SQL 2000 or SQL 2005?
In 2005 is very easily accomplished using Service Broker. Instead of calling
the procedure 500 times, send 500 messages to a service that uses an
activated stored procedure. The message body would contain the arguments you
now pass in. Using the 'max_queue_readers' clause of CREATE QUEUE (see
http://msdn2.microsoft.com/en-us/library/ms190495(en-US,SQL.90).aspx) you
can control the maximum number of procedures to run in paralel.
--
This posting is provided "AS IS" with no warranties, and confers no rights.
HTH,
~ Remus Rusanu
SQL Service Broker
http://msdn2.microsoft.com/en-us/library/ms166043(en-US,SQL.90).aspx
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134587293.323886.256810@.z14g2000cwz.googlegroups.com...
> Hi
> I have a simple stored proc which runs in less than a second.
> This stored proc calculates different values based on parameters
> passed.
> For example sp_name "test 1 ' and
> sp_name 'test 2 ' are totally different.
>
> Now The issue is this stored proc is called more than 500 times in some
> applications.
> So these applications run for 500+ seconds . Although load on sql
> server is almost negligible. Memory is also hugely available.
> I need your suggestion in running these all at the same time.
> In other words all 500 must be started at the same time ..there is no
> need for one procedure to complete and the next to start. None of the
> results are depenedent on any other results.
>
> pls help me .
> Kannan
>|||> So these applications run for 500+ seconds . Although load on sql
> server is almost negligible. Memory is also hugely available.
Remus provided some SQL 2005 suggestions to improve throughput but I suggest
you do root cause performance analysis rather than throw parallel processing
at the problem.
Why do the procs run so long? Disk i/o? CPU? Network utilization? Do you
have SET NOCOUNT ON in your procs? If the load on SQL Server is low and you
have no resource problem, the procs should either run very fast or you
should observe high resource utilization.
--
Hope this helps.
Dan Guzman
SQL Server MVP
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134587293.323886.256810@.z14g2000cwz.googlegroups.com...
> Hi
> I have a simple stored proc which runs in less than a second.
> This stored proc calculates different values based on parameters
> passed.
> For example sp_name "test 1 ' and
> sp_name 'test 2 ' are totally different.
>
> Now The issue is this stored proc is called more than 500 times in some
> applications.
> So these applications run for 500+ seconds . Although load on sql
> server is almost negligible. Memory is also hugely available.
> I need your suggestion in running these all at the same time.
> In other words all 500 must be started at the same time ..there is no
> need for one procedure to complete and the next to start. None of the
> results are depenedent on any other results.
>
> pls help me .
> Kannan
>|||Its SQL server 2000.
Its not a single proc that runs for 500 seconds its becuse the same
procedure is called 500 times with different parameters.
As I mentioned if you run the stored proc once it wont even run for a
second.
There are applications which take input from users and invoke my stored
procedure , The user may enter one value or 500 value.
If the user enters 500 values then the application calls my stored proc
500 times one after another.
That is what is causing a problem , So I created another proc which can
take any number of parameters and then invoke my orignal proc .
Like exec super_proc a,b,c will invoke
exec original_proc 'a'
exec original_proc 'b'
exec original_proc 'c'
Here again SQL server starts executing the next proc after it finishes
first.
I dont want sql server to execute serially but paralelly.
This explains why the load and memory are not utilized fully.
regards
Kannan|||> Its not a single proc that runs for 500 seconds its becuse the same
> procedure is called 500 times with different parameters.
> As I mentioned if you run the stored proc once it wont even run for a
> second.
Yes, I gathered this from your original post. My concern is that one second
is a *very* long time. I have seen single-threaded applications repeatedly
execute an insert stored procedure several hundred times per second. If
each of your proc executions run for a second, I would expect that you would
see high resource utilization (CPU, disk or network) for the duration of the
application run. If you don't, then something is wrong. You can certainly
throw additional threads at the problem but I suspect the performance
improvement won't be as good as you would like. For example, if you are I/O
bound, parallel processing might actually hurt performance.
> If the user enters 500 values then the application calls my stored proc
> 500 times one after another.
> That is what is causing a problem , So I created another proc which can
> take any number of parameters and then invoke my orignal proc .
This might make a miniscule difference due to reduced latency but it's
obviously not the long pole in the tent. I don't think you are addressing
the root cause of your slowness. Can you provide more details on the
calculations the proc performs. Perhaps some code and DDL would help.
--
Hope this helps.
Dan Guzman
SQL Server MVP
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134591179.063554.66800@.g14g2000cwa.googlegroups.com...
> Its SQL server 2000.
> Its not a single proc that runs for 500 seconds its becuse the same
> procedure is called 500 times with different parameters.
> As I mentioned if you run the stored proc once it wont even run for a
> second.
> There are applications which take input from users and invoke my stored
> procedure , The user may enter one value or 500 value.
> If the user enters 500 values then the application calls my stored proc
> 500 times one after another.
> That is what is causing a problem , So I created another proc which can
> take any number of parameters and then invoke my orignal proc .
> Like exec super_proc a,b,c will invoke
> exec original_proc 'a'
> exec original_proc 'b'
> exec original_proc 'c'
> Here again SQL server starts executing the next proc after it finishes
> first.
> I dont want sql server to execute serially but paralelly.
> This explains why the load and memory are not utilized fully.
> regards
> Kannan
>|||I used one second just to illustrate my point. It actually runs in
milliseconds as u have mentioned.
I am not sure if i can post the code / Let me check on this .
will opening multiple connections from end application help ' multiple
connections will be run without any wait time right.
In other words if I run the same proc from two places at the same time
then the second proc will not wait for first one to complete right ?
This is what I want to achive using one connection !.
Kannan|||>I used one second just to illustrate my point. It actually runs in
> milliseconds as u have mentioned.
Is the 500 seconds total elapsed time accurate? Total elapsed time should
be roughly the sum the proc duration plus application processing time. So,
if you run a proc 500 times at an average of 10 ms. each, total duration
ought to be about 5 seconds.
> will opening multiple connections from end application help ' multiple
> connections will be run without any wait time right.
> In other words if I run the same proc from two places at the same time
> then the second proc will not wait for first one to complete right ?
> This is what I want to achive using one connection !.
There should be very little 'wait' time between proc executions on a single
connection. Perhaps a little latency but with only 500 executions, it
shouldn't add up to more than a few milliseconds total.
Multiple connections (multi-threaded processing) might help if you are not
resource constrained but you really need to know the cause of the problem is
before you develop a solution.
> I am not sure if i can post the code / Let me check on this .
The code might not be needed if the proc only runs in few milliseconds.
There's something else going on if the total elapsed time is high.
--
Hope this helps.
Dan Guzman
SQL Server MVP
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134594222.338716.212980@.g43g2000cwa.googlegroups.com...
>I used one second just to illustrate my point. It actually runs in
> milliseconds as u have mentioned.
> I am not sure if i can post the code / Let me check on this .
> will opening multiple connections from end application help ' multiple
> connections will be run without any wait time right.
> In other words if I run the same proc from two places at the same time
> then the second proc will not wait for first one to complete right ?
> This is what I want to achive using one connection !.
> Kannan
>

Run two stored procs in parallel

Hi
I have a simple stored proc which runs in less than a second.
This stored proc calculates different values based on parameters
passed.
For example sp_name "test 1 ' and
sp_name 'test 2 ' are totally different.
Now The issue is this stored proc is called more than 500 times in some
applications.
So these applications run for 500+ seconds . Although load on sql
server is almost negligible. Memory is also hugely available.
I need your suggestion in running these all at the same time.
In other words all 500 must be started at the same time ..there is no
need for one procedure to complete and the next to start. None of the
results are depenedent on any other results.
pls help me .
Kannan
Is this SQL 2000 or SQL 2005?
In 2005 is very easily accomplished using Service Broker. Instead of calling
the procedure 500 times, send 500 messages to a service that uses an
activated stored procedure. The message body would contain the arguments you
now pass in. Using the 'max_queue_readers' clause of CREATE QUEUE (see
http://msdn2.microsoft.com/en-us/library/ms190495(en-US,SQL.90).aspx) you
can control the maximum number of procedures to run in paralel.
This posting is provided "AS IS" with no warranties, and confers no rights.
HTH,
~ Remus Rusanu
SQL Service Broker
http://msdn2.microsoft.com/en-us/library/ms166043(en-US,SQL.90).aspx
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134587293.323886.256810@.z14g2000cwz.googlegr oups.com...
> Hi
> I have a simple stored proc which runs in less than a second.
> This stored proc calculates different values based on parameters
> passed.
> For example sp_name "test 1 ' and
> sp_name 'test 2 ' are totally different.
>
> Now The issue is this stored proc is called more than 500 times in some
> applications.
> So these applications run for 500+ seconds . Although load on sql
> server is almost negligible. Memory is also hugely available.
> I need your suggestion in running these all at the same time.
> In other words all 500 must be started at the same time ..there is no
> need for one procedure to complete and the next to start. None of the
> results are depenedent on any other results.
>
> pls help me .
> Kannan
>
|||> So these applications run for 500+ seconds . Although load on sql
> server is almost negligible. Memory is also hugely available.
Remus provided some SQL 2005 suggestions to improve throughput but I suggest
you do root cause performance analysis rather than throw parallel processing
at the problem.
Why do the procs run so long? Disk i/o? CPU? Network utilization? Do you
have SET NOCOUNT ON in your procs? If the load on SQL Server is low and you
have no resource problem, the procs should either run very fast or you
should observe high resource utilization.
Hope this helps.
Dan Guzman
SQL Server MVP
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134587293.323886.256810@.z14g2000cwz.googlegr oups.com...
> Hi
> I have a simple stored proc which runs in less than a second.
> This stored proc calculates different values based on parameters
> passed.
> For example sp_name "test 1 ' and
> sp_name 'test 2 ' are totally different.
>
> Now The issue is this stored proc is called more than 500 times in some
> applications.
> So these applications run for 500+ seconds . Although load on sql
> server is almost negligible. Memory is also hugely available.
> I need your suggestion in running these all at the same time.
> In other words all 500 must be started at the same time ..there is no
> need for one procedure to complete and the next to start. None of the
> results are depenedent on any other results.
>
> pls help me .
> Kannan
>
|||Its SQL server 2000.
Its not a single proc that runs for 500 seconds its becuse the same
procedure is called 500 times with different parameters.
As I mentioned if you run the stored proc once it wont even run for a
second.
There are applications which take input from users and invoke my stored
procedure , The user may enter one value or 500 value.
If the user enters 500 values then the application calls my stored proc
500 times one after another.
That is what is causing a problem , So I created another proc which can
take any number of parameters and then invoke my orignal proc .
Like exec super_proc a,b,c will invoke
exec original_proc 'a'
exec original_proc 'b'
exec original_proc 'c'
Here again SQL server starts executing the next proc after it finishes
first.
I dont want sql server to execute serially but paralelly.
This explains why the load and memory are not utilized fully.
regards
Kannan
|||> Its not a single proc that runs for 500 seconds its becuse the same
> procedure is called 500 times with different parameters.
> As I mentioned if you run the stored proc once it wont even run for a
> second.
Yes, I gathered this from your original post. My concern is that one second
is a *very* long time. I have seen single-threaded applications repeatedly
execute an insert stored procedure several hundred times per second. If
each of your proc executions run for a second, I would expect that you would
see high resource utilization (CPU, disk or network) for the duration of the
application run. If you don't, then something is wrong. You can certainly
throw additional threads at the problem but I suspect the performance
improvement won't be as good as you would like. For example, if you are I/O
bound, parallel processing might actually hurt performance.

> If the user enters 500 values then the application calls my stored proc
> 500 times one after another.
> That is what is causing a problem , So I created another proc which can
> take any number of parameters and then invoke my orignal proc .
This might make a miniscule difference due to reduced latency but it's
obviously not the long pole in the tent. I don't think you are addressing
the root cause of your slowness. Can you provide more details on the
calculations the proc performs. Perhaps some code and DDL would help.
Hope this helps.
Dan Guzman
SQL Server MVP
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134591179.063554.66800@.g14g2000cwa.googlegro ups.com...
> Its SQL server 2000.
> Its not a single proc that runs for 500 seconds its becuse the same
> procedure is called 500 times with different parameters.
> As I mentioned if you run the stored proc once it wont even run for a
> second.
> There are applications which take input from users and invoke my stored
> procedure , The user may enter one value or 500 value.
> If the user enters 500 values then the application calls my stored proc
> 500 times one after another.
> That is what is causing a problem , So I created another proc which can
> take any number of parameters and then invoke my orignal proc .
> Like exec super_proc a,b,c will invoke
> exec original_proc 'a'
> exec original_proc 'b'
> exec original_proc 'c'
> Here again SQL server starts executing the next proc after it finishes
> first.
> I dont want sql server to execute serially but paralelly.
> This explains why the load and memory are not utilized fully.
> regards
> Kannan
>
|||I used one second just to illustrate my point. It actually runs in
milliseconds as u have mentioned.
I am not sure if i can post the code / Let me check on this .
will opening multiple connections from end application help ? multiple
connections will be run without any wait time right.
In other words if I run the same proc from two places at the same time
then the second proc will not wait for first one to complete right ?
This is what I want to achive using one connection !.
Kannan
|||>I used one second just to illustrate my point. It actually runs in
> milliseconds as u have mentioned.
Is the 500 seconds total elapsed time accurate? Total elapsed time should
be roughly the sum the proc duration plus application processing time. So,
if you run a proc 500 times at an average of 10 ms. each, total duration
ought to be about 5 seconds.

> will opening multiple connections from end application help ? multiple
> connections will be run without any wait time right.
> In other words if I run the same proc from two places at the same time
> then the second proc will not wait for first one to complete right ?
> This is what I want to achive using one connection !.
There should be very little 'wait' time between proc executions on a single
connection. Perhaps a little latency but with only 500 executions, it
shouldn't add up to more than a few milliseconds total.
Multiple connections (multi-threaded processing) might help if you are not
resource constrained but you really need to know the cause of the problem is
before you develop a solution.

> I am not sure if i can post the code / Let me check on this .
The code might not be needed if the proc only runs in few milliseconds.
There's something else going on if the total elapsed time is high.
Hope this helps.
Dan Guzman
SQL Server MVP
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134594222.338716.212980@.g43g2000cwa.googlegr oups.com...
>I used one second just to illustrate my point. It actually runs in
> milliseconds as u have mentioned.
> I am not sure if i can post the code / Let me check on this .
> will opening multiple connections from end application help ? multiple
> connections will be run without any wait time right.
> In other words if I run the same proc from two places at the same time
> then the second proc will not wait for first one to complete right ?
> This is what I want to achive using one connection !.
> Kannan
>

Run two stored procs in parallel

Hi
I have a simple stored proc which runs in less than a second.
This stored proc calculates different values based on parameters
passed.
For example sp_name "test 1 ' and
sp_name 'test 2 ' are totally different.
Now The issue is this stored proc is called more than 500 times in some
applications.
So these applications run for 500+ seconds . Although load on sql
server is almost negligible. Memory is also hugely available.
I need your suggestion in running these all at the same time.
In other words all 500 must be started at the same time ..there is no
need for one procedure to complete and the next to start. None of the
results are depenedent on any other results.
pls help me .
KannanIs this SQL 2000 or SQL 2005?
In 2005 is very easily accomplished using Service Broker. Instead of calling
the procedure 500 times, send 500 messages to a service that uses an
activated stored procedure. The message body would contain the arguments you
now pass in. Using the 'max_queue_readers' clause of CREATE QUEUE (see
http://msdn2.microsoft.com/en-us/library/ms190495(en-US,SQL.90).aspx) you
can control the maximum number of procedures to run in paralel.
This posting is provided "AS IS" with no warranties, and confers no rights.
HTH,
~ Remus Rusanu
SQL Service Broker
http://msdn2.microsoft.com/en-us/library/ms166043(en-US,SQL.90).aspx
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134587293.323886.256810@.z14g2000cwz.googlegroups.com...
> Hi
> I have a simple stored proc which runs in less than a second.
> This stored proc calculates different values based on parameters
> passed.
> For example sp_name "test 1 ' and
> sp_name 'test 2 ' are totally different.
>
> Now The issue is this stored proc is called more than 500 times in some
> applications.
> So these applications run for 500+ seconds . Although load on sql
> server is almost negligible. Memory is also hugely available.
> I need your suggestion in running these all at the same time.
> In other words all 500 must be started at the same time ..there is no
> need for one procedure to complete and the next to start. None of the
> results are depenedent on any other results.
>
> pls help me .
> Kannan
>|||> So these applications run for 500+ seconds . Although load on sql
> server is almost negligible. Memory is also hugely available.
Remus provided some SQL 2005 suggestions to improve throughput but I suggest
you do root cause performance analysis rather than throw parallel processing
at the problem.
Why do the procs run so long? Disk i/o? CPU? Network utilization? Do you
have SET NOCOUNT ON in your procs? If the load on SQL Server is low and you
have no resource problem, the procs should either run very fast or you
should observe high resource utilization.
Hope this helps.
Dan Guzman
SQL Server MVP
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134587293.323886.256810@.z14g2000cwz.googlegroups.com...
> Hi
> I have a simple stored proc which runs in less than a second.
> This stored proc calculates different values based on parameters
> passed.
> For example sp_name "test 1 ' and
> sp_name 'test 2 ' are totally different.
>
> Now The issue is this stored proc is called more than 500 times in some
> applications.
> So these applications run for 500+ seconds . Although load on sql
> server is almost negligible. Memory is also hugely available.
> I need your suggestion in running these all at the same time.
> In other words all 500 must be started at the same time ..there is no
> need for one procedure to complete and the next to start. None of the
> results are depenedent on any other results.
>
> pls help me .
> Kannan
>|||Its SQL server 2000.
Its not a single proc that runs for 500 seconds its becuse the same
procedure is called 500 times with different parameters.
As I mentioned if you run the stored proc once it wont even run for a
second.
There are applications which take input from users and invoke my stored
procedure , The user may enter one value or 500 value.
If the user enters 500 values then the application calls my stored proc
500 times one after another.
That is what is causing a problem , So I created another proc which can
take any number of parameters and then invoke my orignal proc .
Like exec super_proc a,b,c will invoke
exec original_proc 'a'
exec original_proc 'b'
exec original_proc 'c'
Here again SQL server starts executing the next proc after it finishes
first.
I dont want sql server to execute serially but paralelly.
This explains why the load and memory are not utilized fully.
regards
Kannan|||> Its not a single proc that runs for 500 seconds its becuse the same
> procedure is called 500 times with different parameters.
> As I mentioned if you run the stored proc once it wont even run for a
> second.
Yes, I gathered this from your original post. My concern is that one second
is a *very* long time. I have seen single-threaded applications repeatedly
execute an insert stored procedure several hundred times per second. If
each of your proc executions run for a second, I would expect that you would
see high resource utilization (CPU, disk or network) for the duration of the
application run. If you don't, then something is wrong. You can certainly
throw additional threads at the problem but I suspect the performance
improvement won't be as good as you would like. For example, if you are I/O
bound, parallel processing might actually hurt performance.

> If the user enters 500 values then the application calls my stored proc
> 500 times one after another.
> That is what is causing a problem , So I created another proc which can
> take any number of parameters and then invoke my orignal proc .
This might make a miniscule difference due to reduced latency but it's
obviously not the long pole in the tent. I don't think you are addressing
the root cause of your slowness. Can you provide more details on the
calculations the proc performs. Perhaps some code and DDL would help.
Hope this helps.
Dan Guzman
SQL Server MVP
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134591179.063554.66800@.g14g2000cwa.googlegroups.com...
> Its SQL server 2000.
> Its not a single proc that runs for 500 seconds its becuse the same
> procedure is called 500 times with different parameters.
> As I mentioned if you run the stored proc once it wont even run for a
> second.
> There are applications which take input from users and invoke my stored
> procedure , The user may enter one value or 500 value.
> If the user enters 500 values then the application calls my stored proc
> 500 times one after another.
> That is what is causing a problem , So I created another proc which can
> take any number of parameters and then invoke my orignal proc .
> Like exec super_proc a,b,c will invoke
> exec original_proc 'a'
> exec original_proc 'b'
> exec original_proc 'c'
> Here again SQL server starts executing the next proc after it finishes
> first.
> I dont want sql server to execute serially but paralelly.
> This explains why the load and memory are not utilized fully.
> regards
> Kannan
>|||I used one second just to illustrate my point. It actually runs in
milliseconds as u have mentioned.
I am not sure if i can post the code / Let me check on this .
will opening multiple connections from end application help ' multiple
connections will be run without any wait time right.
In other words if I run the same proc from two places at the same time
then the second proc will not wait for first one to complete right ?
This is what I want to achive using one connection !.
Kannan|||>I used one second just to illustrate my point. It actually runs in
> milliseconds as u have mentioned.
Is the 500 seconds total elapsed time accurate? Total elapsed time should
be roughly the sum the proc duration plus application processing time. So,
if you run a proc 500 times at an average of 10 ms. each, total duration
ought to be about 5 seconds.

> will opening multiple connections from end application help ' multiple
> connections will be run without any wait time right.
> In other words if I run the same proc from two places at the same time
> then the second proc will not wait for first one to complete right ?
> This is what I want to achive using one connection !.
There should be very little 'wait' time between proc executions on a single
connection. Perhaps a little latency but with only 500 executions, it
shouldn't add up to more than a few milliseconds total.
Multiple connections (multi-threaded processing) might help if you are not
resource constrained but you really need to know the cause of the problem is
before you develop a solution.

> I am not sure if i can post the code / Let me check on this .
The code might not be needed if the proc only runs in few milliseconds.
There's something else going on if the total elapsed time is high.
Hope this helps.
Dan Guzman
SQL Server MVP
"Kannan" <Kannanmca@.gmail.com> wrote in message
news:1134594222.338716.212980@.g43g2000cwa.googlegroups.com...
>I used one second just to illustrate my point. It actually runs in
> milliseconds as u have mentioned.
> I am not sure if i can post the code / Let me check on this .
> will opening multiple connections from end application help ' multiple
> connections will be run without any wait time right.
> In other words if I run the same proc from two places at the same time
> then the second proc will not wait for first one to complete right ?
> This is what I want to achive using one connection !.
> Kannan
>

RUN "Intergration Services Deployment Manifest" through Command Prompt

Hi, I want to run "Intergration Services Deployment Manifest" through the command prompt, passing the values .. which should basically go and install that Package either on File System or SQL Server. Does any one have any idea on how to do that ...

Thanks,

You can use the dtutil command prompt utility to copy a package to the Microsoft SQL Server database, the SSIS Package Store, or the file system.

However, the package must already exist in one of these three locations.

For more information, see "dtutil Utility" at http://msdn2.microsoft.com/en-us/library/ms162820.aspx.

|||Thanks alot for your reply ..

RUN "Intergration Services Deployment Manifest" through Command Prompt

Hi, I want to run "Intergration Services Deployment Manifest" through the command prompt, passing the values .. which should basically go and install that Package either on File System or SQL Server. Does any one have any idea on how to do that ...

Thanks,

You can use the dtutil command prompt utility to copy a package to the Microsoft SQL Server database, the SSIS Package Store, or the file system.

However, the package must already exist in one of these three locations.

For more information, see "dtutil Utility" at http://msdn2.microsoft.com/en-us/library/ms162820.aspx.

|||Thanks alot for your reply ..