Hi,
I have a report with multiple pages.
I added a page break but then went to the table properties and unchecked the
page break checkbox.
However, when i run the report from my vb.net application I can see a page
break at the end of each page. How can I delete the page breaks?
Thanks.You can get page breaks from groups, lists, etc... Either search through
the RDL or check each one of the items to ensure your not asking for page
breaks somewhere else...
--
Wayne Snyder, MCDBA, SQL Server MVP
Mariner, Charlotte, NC
www.mariner-usa.com
(Please respond only to the newsgroups.)
I support the Professional Association of SQL Server (PASS) and it's
community of SQL Server professionals.
www.sqlpass.org
"collie" <collie@.discussions.microsoft.com> wrote in message
news:ABEEBB6A-A408-48C0-A528-F881D5815A65@.microsoft.com...
> Hi,
> I have a report with multiple pages.
> I added a page break but then went to the table properties and unchecked
the
> page break checkbox.
> However, when i run the report from my vb.net application I can see a page
> break at the end of each page. How can I delete the page breaks?
> Thanks.
Showing posts with label multiple. Show all posts
Showing posts with label multiple. Show all posts
Friday, March 30, 2012
Friday, March 23, 2012
Package loses configuration files!
I created a project with multiple packages. I created a global config file that all packages reference. I also created a config file for each package that contains the package specific settings. I checked it all into TFS. I went to a different machine and checked it all out. Upon loading a package into the IDE I got the following 3 warnings.
Warning loading MT_LSE_PROD_StageLoad.dtsx: The configuration file "MT_LSE_PROD_StageLoad.dtsConfig" cannot be found. Check the directory and file name. e:\contentloader\sprint1a\MT_LSE_PROD_StageLoad.dtsx
Warning loading MT_LSE_PROD_StageLoad.dtsx: The configuration file "Environment.dtsConfig" cannot be found. Check the directory and file name. e:\contentloader\sprint1a\MT_LSE_PROD_StageLoad.dtsx
Warning loading MT_LSE_PROD_StageLoad.dtsx: Failed to load at least one of the configuration entries for the package. Check configurations entries and previous warnings to see descriptions of which configuration failed. e:\contentloader\sprint1a\MT_LSE_PROD_StageLoad.dtsx
Any ideas? And yes, the files *ARE* there! I tried building the project and it builds. I also tried doing a deployment build and that too built. I am so confused!!!
Tuesday, March 20, 2012
Package error: Cannot create thread
I have a child package that has been run successfully multiple times in the last month +. Each time with roughly the same amount of data, give or take a few thousand rows.
Suddenly, this child package is now giving me the following errors from the log file:
Error: 2006-11-17 12:04:19.98
Code: 0xC0047031
Source: DFLT Primary DTS.Pipeline
Description: The Data Flow task failed to create a required thread and cannot begin running. The usually occurs when there is an out-of-memory state.
End Error
Error: 2006-11-17 12:04:20.03
Code: 0xC004700E
Source: DFLT Primary DTS.Pipeline
Description: The Data Flow task engine failed at startup because it cannot create one or more required threads.
End Error
I tried taking the child out of parent and running it by itself. I still get the same error. There are three other child packages that run on the exact same data and they have no problems. The control flow for the package first runs an SQL command. Then it has a data flow. The data flow grabs records from the source, adds two derived columns, looks up data and then stores to the destination. Relatively easy compared to other packages that are running just fine.
I've had our network people check the both the server running SSIS and the database server (two different machines) and there are no memory spike while the package is running.
Any ideas?
The problem could also be related with all the other applications/processes running on that box. Can you try stopping other SSIS packages and applications, and try running the problematic one? if that works, that means the machine is maxed out on available number of threads. or memory...
It's also possible that other applications might be running zombie processes on the box, even if the process seems to exit, there could be a thread or memory leak.
I'd use perfmon tool to read some of the critical resources for the box in which SSIS is running. You don't need to look at the database server, as SSIS creates a new process only on the box it's running. I'd check : thread/process, total thread, memory/process, total memory, page faults.
Package Configurations - Using One config file for Multiple Packages
Hi All,
I am working on a project currently where we have many SSIS packages and we want to minimize the number of config files to be used. What I was thinking was using one master config file which will have Server/Database info which will be used by all the packages. Now the thing is there are multiple packages which are being executed from within a master package and I was wondering if this will be an issue. Is there anything else that I will have to keep in mind? Any help is appreciated.
Thanks
If my memory serves correctly then you will get lots of warnings about references in your config file to properties that don't exist in your package. But it should still execute OK.
-Jamie
|||Thanks JamieP2P reindexing
I was looking at setting up multiple SQL Server 2005 Standard edition
servers using Peer-to-Peer Replication with the goal being able to
reindex one server at a time so as not to have any downtime (via locked
tables). Would this work?
Also, I know log-shipping to a standby server only requires 1 CPU
license which is attractive as CPU licenses are pricey. Still, if I
use P2P only for this reindex purpose (including three nodes for
voting) with only one server being used by customers/users, is a 1 CPU
license good enough?
Thanks!
MichaelI can't seem to find an answer anywhere for these questions:
Can I run ALTER INDEX REBUILD seperately between a live SQL 2005 server
and the hot-standby server? This way, if the live table is locked
during this operation, the hot-standby server would take over the CRUD
statements.
Is it allowed to have a single CPU license for two SQL 2005 server if
one is a hot-standby-only server and the standby-only server is only
used for failover purposes. Note that when tables are locked during
ALTER INDEX REBUILD, I assume the failover would kick in. Is that
possible?
The goal is to have no downtime for executing ALTER INDEX REBUILD.
Thank you,
Michael
servers using Peer-to-Peer Replication with the goal being able to
reindex one server at a time so as not to have any downtime (via locked
tables). Would this work?
Also, I know log-shipping to a standby server only requires 1 CPU
license which is attractive as CPU licenses are pricey. Still, if I
use P2P only for this reindex purpose (including three nodes for
voting) with only one server being used by customers/users, is a 1 CPU
license good enough?
Thanks!
MichaelI can't seem to find an answer anywhere for these questions:
Can I run ALTER INDEX REBUILD seperately between a live SQL 2005 server
and the hot-standby server? This way, if the live table is locked
during this operation, the hot-standby server would take over the CRUD
statements.
Is it allowed to have a single CPU license for two SQL 2005 server if
one is a hot-standby-only server and the standby-only server is only
used for failover purposes. Note that when tables are locked during
ALTER INDEX REBUILD, I assume the failover would kick in. Is that
possible?
The goal is to have no downtime for executing ALTER INDEX REBUILD.
Thank you,
Michael
Monday, March 12, 2012
P2P reindexing
I was looking at setting up multiple SQL Server 2005 Standard edition
servers using Peer-to-Peer Replication with the goal being able to
reindex one server at a time so as not to have any downtime (via locked
tables). Would this work?
Also, I know log-shipping to a standby server only requires 1 CPU
license which is attractive as CPU licenses are pricey. Still, if I
use P2P only for this reindex purpose (including three nodes for
voting) with only one server being used by customers/users, is a 1 CPU
license good enough?
Thanks!
MichaelI can't seem to find an answer anywhere for these questions:
Can I run ALTER INDEX REBUILD seperately between a live SQL 2005 server
and the hot-standby server? This way, if the live table is locked
during this operation, the hot-standby server would take over the CRUD
statements.
Is it allowed to have a single CPU license for two SQL 2005 server if
one is a hot-standby-only server and the standby-only server is only
used for failover purposes. Note that when tables are locked during
ALTER INDEX REBUILD, I assume the failover would kick in. Is that
possible?
The goal is to have no downtime for executing ALTER INDEX REBUILD.
Thank you,
Michael
servers using Peer-to-Peer Replication with the goal being able to
reindex one server at a time so as not to have any downtime (via locked
tables). Would this work?
Also, I know log-shipping to a standby server only requires 1 CPU
license which is attractive as CPU licenses are pricey. Still, if I
use P2P only for this reindex purpose (including three nodes for
voting) with only one server being used by customers/users, is a 1 CPU
license good enough?
Thanks!
MichaelI can't seem to find an answer anywhere for these questions:
Can I run ALTER INDEX REBUILD seperately between a live SQL 2005 server
and the hot-standby server? This way, if the live table is locked
during this operation, the hot-standby server would take over the CRUD
statements.
Is it allowed to have a single CPU license for two SQL 2005 server if
one is a hot-standby-only server and the standby-only server is only
used for failover purposes. Note that when tables are locked during
ALTER INDEX REBUILD, I assume the failover would kick in. Is that
possible?
The goal is to have no downtime for executing ALTER INDEX REBUILD.
Thank you,
Michael
Friday, March 9, 2012
P2P reindexing
I was looking at setting up multiple SQL Server 2005 Standard edition
servers using Peer-to-Peer Replication with the goal being able to
reindex one server at a time so as not to have any downtime (via locked
tables). Would this work?
Also, I know log-shipping to a standby server only requires 1 CPU
license which is attractive as CPU licenses are pricey. Still, if I
use P2P only for this reindex purpose (including three nodes for
voting) with only one server being used by customers/users, is a 1 CPU
license good enough?
Thanks!
MichaelI can't seem to find an answer anywhere for these questions:
Can I run ALTER INDEX REBUILD seperately between a live SQL 2005 server
and the hot-standby server? This way, if the live table is locked
during this operation, the hot-standby server would take over the CRUD
statements.
Is it allowed to have a single CPU license for two SQL 2005 server if
one is a hot-standby-only server and the standby-only server is only
used for failover purposes. Note that when tables are locked during
ALTER INDEX REBUILD, I assume the failover would kick in. Is that
possible?
The goal is to have no downtime for executing ALTER INDEX REBUILD.
Thank you,
Michael
servers using Peer-to-Peer Replication with the goal being able to
reindex one server at a time so as not to have any downtime (via locked
tables). Would this work?
Also, I know log-shipping to a standby server only requires 1 CPU
license which is attractive as CPU licenses are pricey. Still, if I
use P2P only for this reindex purpose (including three nodes for
voting) with only one server being used by customers/users, is a 1 CPU
license good enough?
Thanks!
MichaelI can't seem to find an answer anywhere for these questions:
Can I run ALTER INDEX REBUILD seperately between a live SQL 2005 server
and the hot-standby server? This way, if the live table is locked
during this operation, the hot-standby server would take over the CRUD
statements.
Is it allowed to have a single CPU license for two SQL 2005 server if
one is a hot-standby-only server and the standby-only server is only
used for failover purposes. Note that when tables are locked during
ALTER INDEX REBUILD, I assume the failover would kick in. Is that
possible?
The goal is to have no downtime for executing ALTER INDEX REBUILD.
Thank you,
Michael
Monday, February 20, 2012
Overhead of Multiple SQL Calls
Hey folks,
I've got a client server C# Windows Form app that communicates with a SQL
Server database backend. The number of users is less than 50.
I'm wondering what the difference in overhead is when say:
a) A Stored Procedure is called ten times, returning 1 record each time
VS.
b) The same Stored Procedure is called once returning 10 records
Due to connection pooling I would think the overhead is somewhat negligible,
especially when there's less than 50 users. I know option 'b' would be the
preffered method when it comes to database communication, but is there
really a BIG difference between the two when dealing with this few users?
Thanks!
(also posted on windows.forms newsgroup on Feb 25 with no response)Hi John,
I would think the network traffic would be significantly greater when
submiting 10 calls to get 1 record each as opposed to 1 call to get 10
records. That's automatically 10X the client-to-server network traffic.
Constantly opening and closing connections, as well as maintaining
connections that aren't in use, can use up a lot of resources on SQL Server
also. Now whether this should cause problems with a small number of users
is another story... It probably depends on how they're using the server...
Are they running automated reports that are requesting records one at a time
and might include thousands of separate requests? Or maybe they're just
accessing 30 records/hour on average each?
Connection pooling helps, but it can't compensate completely for a design
flaw.
Are you using C#.NET? If so, you might look into ADO.NET. ADO.NET is
designed to handle a lot of the management issues for you.
Thanks,
Mike C.
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>|||You have that many today but what about tomorrow? Seriously you should
always keep scalability in mind when designing the app and the schema. In
your example you are incurring about 10 times the overhead on not only the
server but the network and the client. There certainly may be times when it
might make sense to do it the A way but if you can do it the B way then you
should.
--
Andrew J. Kelly SQL MVP
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>
I've got a client server C# Windows Form app that communicates with a SQL
Server database backend. The number of users is less than 50.
I'm wondering what the difference in overhead is when say:
a) A Stored Procedure is called ten times, returning 1 record each time
VS.
b) The same Stored Procedure is called once returning 10 records
Due to connection pooling I would think the overhead is somewhat negligible,
especially when there's less than 50 users. I know option 'b' would be the
preffered method when it comes to database communication, but is there
really a BIG difference between the two when dealing with this few users?
Thanks!
(also posted on windows.forms newsgroup on Feb 25 with no response)Hi John,
I would think the network traffic would be significantly greater when
submiting 10 calls to get 1 record each as opposed to 1 call to get 10
records. That's automatically 10X the client-to-server network traffic.
Constantly opening and closing connections, as well as maintaining
connections that aren't in use, can use up a lot of resources on SQL Server
also. Now whether this should cause problems with a small number of users
is another story... It probably depends on how they're using the server...
Are they running automated reports that are requesting records one at a time
and might include thousands of separate requests? Or maybe they're just
accessing 30 records/hour on average each?
Connection pooling helps, but it can't compensate completely for a design
flaw.
Are you using C#.NET? If so, you might look into ADO.NET. ADO.NET is
designed to handle a lot of the management issues for you.
Thanks,
Mike C.
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>|||You have that many today but what about tomorrow? Seriously you should
always keep scalability in mind when designing the app and the schema. In
your example you are incurring about 10 times the overhead on not only the
server but the network and the client. There certainly may be times when it
might make sense to do it the A way but if you can do it the B way then you
should.
--
Andrew J. Kelly SQL MVP
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>
Overhead of Multiple SQL Calls
Hey folks,
I've got a client server C# Windows Form app that communicates with a SQL
Server database backend. The number of users is less than 50.
I'm wondering what the difference in overhead is when say:
a) A Stored Procedure is called ten times, returning 1 record each time
VS.
b) The same Stored Procedure is called once returning 10 records
Due to connection pooling I would think the overhead is somewhat negligible,
especially when there's less than 50 users. I know option 'b' would be the
preffered method when it comes to database communication, but is there
really a BIG difference between the two when dealing with this few users?
Thanks!
(also posted on windows.forms newsgroup on Feb 25 with no response)Hi John,
I would think the network traffic would be significantly greater when
submiting 10 calls to get 1 record each as opposed to 1 call to get 10
records. That's automatically 10X the client-to-server network traffic.
Constantly opening and closing connections, as well as maintaining
connections that aren't in use, can use up a lot of resources on SQL Server
also. Now whether this should cause problems with a small number of users
is another story... It probably depends on how they're using the server...
Are they running automated reports that are requesting records one at a time
and might include thousands of separate requests? Or maybe they're just
accessing 30 records/hour on average each?
Connection pooling helps, but it can't compensate completely for a design
flaw.
Are you using C#.NET? If so, you might look into ADO.NET. ADO.NET is
designed to handle a lot of the management issues for you.
Thanks,
Mike C.
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>|||You have that many today but what about tomorrow? Seriously you should
always keep scalability in mind when designing the app and the schema. In
your example you are incurring about 10 times the overhead on not only the
server but the network and the client. There certainly may be times when it
might make sense to do it the A way but if you can do it the B way then you
should.
Andrew J. Kelly SQL MVP
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>
I've got a client server C# Windows Form app that communicates with a SQL
Server database backend. The number of users is less than 50.
I'm wondering what the difference in overhead is when say:
a) A Stored Procedure is called ten times, returning 1 record each time
VS.
b) The same Stored Procedure is called once returning 10 records
Due to connection pooling I would think the overhead is somewhat negligible,
especially when there's less than 50 users. I know option 'b' would be the
preffered method when it comes to database communication, but is there
really a BIG difference between the two when dealing with this few users?
Thanks!
(also posted on windows.forms newsgroup on Feb 25 with no response)Hi John,
I would think the network traffic would be significantly greater when
submiting 10 calls to get 1 record each as opposed to 1 call to get 10
records. That's automatically 10X the client-to-server network traffic.
Constantly opening and closing connections, as well as maintaining
connections that aren't in use, can use up a lot of resources on SQL Server
also. Now whether this should cause problems with a small number of users
is another story... It probably depends on how they're using the server...
Are they running automated reports that are requesting records one at a time
and might include thousands of separate requests? Or maybe they're just
accessing 30 records/hour on average each?
Connection pooling helps, but it can't compensate completely for a design
flaw.
Are you using C#.NET? If so, you might look into ADO.NET. ADO.NET is
designed to handle a lot of the management issues for you.
Thanks,
Mike C.
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>|||You have that many today but what about tomorrow? Seriously you should
always keep scalability in mind when designing the app and the schema. In
your example you are incurring about 10 times the overhead on not only the
server but the network and the client. There certainly may be times when it
might make sense to do it the A way but if you can do it the B way then you
should.
Andrew J. Kelly SQL MVP
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>
Overhead of Multiple SQL Calls
Hey folks,
I've got a client server C# Windows Form app that communicates with a SQL
Server database backend. The number of users is less than 50.
I'm wondering what the difference in overhead is when say:
a) A Stored Procedure is called ten times, returning 1 record each time
VS.
b) The same Stored Procedure is called once returning 10 records
Due to connection pooling I would think the overhead is somewhat negligible,
especially when there's less than 50 users. I know option 'b' would be the
preffered method when it comes to database communication, but is there
really a BIG difference between the two when dealing with this few users?
Thanks!
(also posted on windows.forms newsgroup on Feb 25 with no response)
Hi John,
I would think the network traffic would be significantly greater when
submiting 10 calls to get 1 record each as opposed to 1 call to get 10
records. That's automatically 10X the client-to-server network traffic.
Constantly opening and closing connections, as well as maintaining
connections that aren't in use, can use up a lot of resources on SQL Server
also. Now whether this should cause problems with a small number of users
is another story... It probably depends on how they're using the server...
Are they running automated reports that are requesting records one at a time
and might include thousands of separate requests? Or maybe they're just
accessing 30 records/hour on average each?
Connection pooling helps, but it can't compensate completely for a design
flaw.
Are you using C#.NET? If so, you might look into ADO.NET. ADO.NET is
designed to handle a lot of the management issues for you.
Thanks,
Mike C.
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>
|||You have that many today but what about tomorrow? Seriously you should
always keep scalability in mind when designing the app and the schema. In
your example you are incurring about 10 times the overhead on not only the
server but the network and the client. There certainly may be times when it
might make sense to do it the A way but if you can do it the B way then you
should.
Andrew J. Kelly SQL MVP
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>
I've got a client server C# Windows Form app that communicates with a SQL
Server database backend. The number of users is less than 50.
I'm wondering what the difference in overhead is when say:
a) A Stored Procedure is called ten times, returning 1 record each time
VS.
b) The same Stored Procedure is called once returning 10 records
Due to connection pooling I would think the overhead is somewhat negligible,
especially when there's less than 50 users. I know option 'b' would be the
preffered method when it comes to database communication, but is there
really a BIG difference between the two when dealing with this few users?
Thanks!
(also posted on windows.forms newsgroup on Feb 25 with no response)
Hi John,
I would think the network traffic would be significantly greater when
submiting 10 calls to get 1 record each as opposed to 1 call to get 10
records. That's automatically 10X the client-to-server network traffic.
Constantly opening and closing connections, as well as maintaining
connections that aren't in use, can use up a lot of resources on SQL Server
also. Now whether this should cause problems with a small number of users
is another story... It probably depends on how they're using the server...
Are they running automated reports that are requesting records one at a time
and might include thousands of separate requests? Or maybe they're just
accessing 30 records/hour on average each?
Connection pooling helps, but it can't compensate completely for a design
flaw.
Are you using C#.NET? If so, you might look into ADO.NET. ADO.NET is
designed to handle a lot of the management issues for you.
Thanks,
Mike C.
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>
|||You have that many today but what about tomorrow? Seriously you should
always keep scalability in mind when designing the app and the schema. In
your example you are incurring about 10 times the overhead on not only the
server but the network and the client. There certainly may be times when it
might make sense to do it the A way but if you can do it the B way then you
should.
Andrew J. Kelly SQL MVP
"John Smith" <js@.no.com> wrote in message
news:OHFsJQqHFHA.560@.TK2MSFTNGP12.phx.gbl...
> Hey folks,
> I've got a client server C# Windows Form app that communicates with a SQL
> Server database backend. The number of users is less than 50.
> I'm wondering what the difference in overhead is when say:
> a) A Stored Procedure is called ten times, returning 1 record each time
> VS.
> b) The same Stored Procedure is called once returning 10 records
> Due to connection pooling I would think the overhead is somewhat
> negligible,
> especially when there's less than 50 users. I know option 'b' would be
> the
> preffered method when it comes to database communication, but is there
> really a BIG difference between the two when dealing with this few users?
> Thanks!
> (also posted on windows.forms newsgroup on Feb 25 with no response)
>
Subscribe to:
Posts (Atom)