Network Error...

Hi All,
I have a windows 2003 server and MS SQL Server 2000 Enterprise edtion
installed in it... From past few days I am the schedule jobs are getting
failed... its giving below error...
[Microsoft][ODBC SQL Server Driver][Named Pipes]ConnectionRead
(WrapperRead()).
Server: Msg 11, Level 16, State 1, Line 0
General network error. Check your network documentation.
Plz advice how to get rid of this...
Thanks in advance.
-Mohit.Refer this KB Article. This talks about similar error message when restoring
database.
http://support.microsoft.com/default.aspx?scid=kb;en-us;827452
It looks like it is MDAC bug and work around is change the SQL Server to use
TCP/IP
Ravi
"Mohit" wrote:
> Hi All,
> I have a windows 2003 server and MS SQL Server 2000 Enterprise edtion
> installed in it... From past few days I am the schedule jobs are getting
> failed... its giving below error...
> [Microsoft][ODBC SQL Server Driver][Named Pipes]ConnectionRead
> (WrapperRead()).
> Server: Msg 11, Level 16, State 1, Line 0
> General network error. Check your network documentation.
>
> Plz advice how to get rid of this...
> Thanks in advance.
> -Mohit.
>

Network Error...

Hi All,
I have a windows 2003 server and MS SQL Server 2000 Enterprise edtion
installed in it... From past few days I am the schedule jobs are getting
failed... its giving below error...
[Microsoft][ODBC SQL Server Driver][Named Pipes]ConnectionRead
(WrapperRead()).
Server: Msg 11, Level 16, State 1, Line 0
General network error. Check your network documentation.
Plz advice how to get rid of this...
Thanks in advance.
-Mohit.
Refer this KB Article. This talks about similar error message when restoring
database.
http://support.microsoft.com/default...b;en-us;827452
It looks like it is MDAC bug and work around is change the SQL Server to use
TCP/IP
Ravi
"Mohit" wrote:

> Hi All,
> I have a windows 2003 server and MS SQL Server 2000 Enterprise edtion
> installed in it... From past few days I am the schedule jobs are getting
> failed... its giving below error...
> [Microsoft][ODBC SQL Server Driver][Named Pipes]ConnectionRead
> (WrapperRead()).
> Server: Msg 11, Level 16, State 1, Line 0
> General network error. Check your network documentation.
>
> Plz advice how to get rid of this...
> Thanks in advance.
> -Mohit.
>

Network Error...

Hi All,
I have a windows 2003 server and MS SQL Server 2000 Enterprise edtion
installed in it... From past few days I am the schedule jobs are getting
failed... its giving below error...
[Microsoft][ODBC SQL Server Driver][Named Pipes]ConnectionRead
(WrapperRead()).
Server: Msg 11, Level 16, State 1, Line 0
General network error. Check your network documentation.
Plz advice how to get rid of this...
Thanks in advance.
-Mohit.Refer this KB Article. This talks about similar error message when restoring
database.
http://support.microsoft.com/defaul...kb;en-us;827452
It looks like it is MDAC bug and work around is change the SQL Server to use
TCP/IP
Ravi
"Mohit" wrote:

> Hi All,
> I have a windows 2003 server and MS SQL Server 2000 Enterprise edtion
> installed in it... From past few days I am the schedule jobs are getting
> failed... its giving below error...
> [Microsoft][ODBC SQL Server Driver][Named Pipes]ConnectionRea
d
> (WrapperRead()).
> Server: Msg 11, Level 16, State 1, Line 0
> General network error. Check your network documentation.
>
> Plz advice how to get rid of this...
> Thanks in advance.
> -Mohit.
>sql

Network error.

Hi,
I am facing a weird network problem. I am running a SQL 2000 SP2
enterprise edition over a Windows 2003 enterprise edition machine.
I am getting the network error messages (see below) while trying to
connect to the server. However if i am connected to the server before
the problem occurs i can continue running queries from the query
analyzer but cannot use the object browser to view the tables/ views
etc.
The problem is temporarily fixed if i restart the sql server service
but reappears after some times.
Seems to me it is is not a network problem otherwise the connection
would have broken.
Error 1:
[ODBC SQL Server Driver][DBNETLIB]General network error. Check your
network documentation.
Error 2:
[ODBC SQL Server Driver][DBNETLIB]ConnectionRead (recv()).
Any idea why this is happening?
Thanks in advance
Taw.
Consider applying SP4, or at least SP3a, to your SQL 2000 instance.
Hope this helps.
Dan Guzman
SQL Server MVP
"Taw" <tawfiq.choudhury@.grameenphone.com> wrote in message
news:1185800793.766706.9030@.l70g2000hse.googlegrou ps.com...
> Hi,
> I am facing a weird network problem. I am running a SQL 2000 SP2
> enterprise edition over a Windows 2003 enterprise edition machine.
> I am getting the network error messages (see below) while trying to
> connect to the server. However if i am connected to the server before
> the problem occurs i can continue running queries from the query
> analyzer but cannot use the object browser to view the tables/ views
> etc.
> The problem is temporarily fixed if i restart the sql server service
> but reappears after some times.
> Seems to me it is is not a network problem otherwise the connection
> would have broken.
> Error 1:
> [ODBC SQL Server Driver][DBNETLIB]General network error. Check your
> network documentation.
> Error 2:
> [ODBC SQL Server Driver][DBNETLIB]ConnectionRead (recv()).
> Any idea why this is happening?
> Thanks in advance
> Taw.
>

Network error.

Hi,
I am facing a weird network problem. I am running a SQL 2000 SP2
enterprise edition over a Windows 2003 enterprise edition machine.
I am getting the network error messages (see below) while trying to
connect to the server. However if i am connected to the server before
the problem occurs i can continue running queries from the query
analyzer but cannot use the object browser to view the tables/ views
etc.
The problem is temporarily fixed if i restart the sql server service
but reappears after some times.
Seems to me it is is not a network problem otherwise the connection
would have broken.
Error 1:
[ODBC SQL Server Driver][DBNETLIB]General network error. Check your
network documentation.
Error 2:
[ODBC SQL Server Driver][DBNETLIB]ConnectionRead (recv()).
Any idea why this is happening?
Thanks in advance
Taw.Consider applying SP4, or at least SP3a, to your SQL 2000 instance.
Hope this helps.
Dan Guzman
SQL Server MVP
"Taw" <tawfiq.choudhury@.grameenphone.com> wrote in message
news:1185800793.766706.9030@.l70g2000hse.googlegroups.com...
> Hi,
> I am facing a weird network problem. I am running a SQL 2000 SP2
> enterprise edition over a Windows 2003 enterprise edition machine.
> I am getting the network error messages (see below) while trying to
> connect to the server. However if i am connected to the server before
> the problem occurs i can continue running queries from the query
> analyzer but cannot use the object browser to view the tables/ views
> etc.
> The problem is temporarily fixed if i restart the sql server service
> but reappears after some times.
> Seems to me it is is not a network problem otherwise the connection
> would have broken.
> Error 1:
> [ODBC SQL Server Driver][DBNETLIB]General network error. Check you
r
> network documentation.
> Error 2:
> [ODBC SQL Server Driver][DBNETLIB]ConnectionRead (recv()).
> Any idea why this is happening?
> Thanks in advance
> Taw.
>

Network error.

Hi,
I am facing a weird network problem. I am running a SQL 2000 SP2
enterprise edition over a Windows 2003 enterprise edition machine.
I am getting the network error messages (see below) while trying to
connect to the server. However if i am connected to the server before
the problem occurs i can continue running queries from the query
analyzer but cannot use the object browser to view the tables/ views
etc.
The problem is temporarily fixed if i restart the sql server service
but reappears after some times.
Seems to me it is is not a network problem otherwise the connection
would have broken.
Error 1:
[ODBC SQL Server Driver][DBNETLIB]General network error. Check your
network documentation.
Error 2:
[ODBC SQL Server Driver][DBNETLIB]ConnectionRead (recv()).
Any idea why this is happening?
Thanks in advance
Taw.Consider applying SP4, or at least SP3a, to your SQL 2000 instance.
--
Hope this helps.
Dan Guzman
SQL Server MVP
"Taw" <tawfiq.choudhury@.grameenphone.com> wrote in message
news:1185800793.766706.9030@.l70g2000hse.googlegroups.com...
> Hi,
> I am facing a weird network problem. I am running a SQL 2000 SP2
> enterprise edition over a Windows 2003 enterprise edition machine.
> I am getting the network error messages (see below) while trying to
> connect to the server. However if i am connected to the server before
> the problem occurs i can continue running queries from the query
> analyzer but cannot use the object browser to view the tables/ views
> etc.
> The problem is temporarily fixed if i restart the sql server service
> but reappears after some times.
> Seems to me it is is not a network problem otherwise the connection
> would have broken.
> Error 1:
> [ODBC SQL Server Driver][DBNETLIB]General network error. Check your
> network documentation.
> Error 2:
> [ODBC SQL Server Driver][DBNETLIB]ConnectionRead (recv()).
> Any idea why this is happening?
> Thanks in advance
> Taw.
>

Netstat shows Port 1433 Activity

When I run Netstat it shows many connections from Port 1433 on my server to
various ports on seemingly random computers. SQL Profiler and Enterprise Man
ager do not show logins or processes that correspond to this activity. Howev
er Task Manager does show increased memory and CPU time for sqlservr.exe As
time goes on the number of netstat connections increases until I reboot the
server.
I have SP3 and also disable Port 1434.
I don't know if this is related but if I Stop and then Start MSSQLSERVER Ser
vice, then SQL doesn't seem to work until I reboot the Server.
Does anyone know what is causing these network connections? I plan to use IP
SEC to filter certain IPs and Ports to reduce this but I would like to under
stand the problem better before I try to solve it.Netstat should show you only valid IP addresses connected to 1433 on your
server.
You can use PortQuery or Port reporter to validate this as well.
832919 New features and functionality in PortQry version 2.0
http://support.microsoft.com/?id=832919
Is SQL Server behind a firewall?
Are you using Standard SQL Security ?
Thanks,
Kevin McDonnell
Microsoft Corporation
This posting is provided AS IS with no warranties, and confers no rights.|||Kevin,
Thanks for your response.
Downloaded Port Reporter. It stated that the server did not provide process
mapping. Downloaded TCPView.
At the time I ran TCPView, instead of the usual connections on port 1433, t
here were many (about 25) established connections on port 1025, some from IP
addresses on the server that should not be establishing connections. All of
these were associated with msdtc.exe. I killed a couple of these processes.
When I killed the 2nd all the remaining disappeared. What does this indicat
e?
When I have observed many connections on port 1433 the state is usually WAIT
_STATE or FIN_2 something.
The server is a leased server at an ISP (Interland) that I administer remote
ly via Terminal Services. There are 2 NICs: 1 to their LAN and 1 to the Inte
rnet.
For firewall I hoped to use IPSEC. I assume that 'Standard SQL Security' mea
ns the same to you as to me (passwords, limit user capabilities) and the ans
wer is Yes.
If the connection state is not "ESTABLISHED' does that mean that other devic
es are attempting to connect unsuccessfully? Could these be port scanner vir
uses or similar?
The msdtc.exe ESTABLISHED connections caused me more concern. Does that mean
my server and another device were 'handshaking' or exchanging data?
Mike

quote:

---

Originally posted by Kevin McDonnell [MSFT]
Netstat should show you only valid IP addresses connected to 1433 on your
server.
You can use PortQuery or Port reporter to validate this as well.
832919 New features and functionality in PortQry version 2.0
http://support.microsoft.com/?id=832919
Is SQL Server behind a firewall?
Are you using Standard SQL Security ?
Thanks,
Kevin McDonnell
Microsoft Corporation
This posting is provided AS IS with no warranties, and confers no rights.

---

|||Previous posting:
At the time I ran TCPView, instead of the usual connections on port
1433, there were many (about 25) established connections on port 1025,
some from IP addresses on the server that should not be establishing
connections. All of these were associated with msdtc.exe. I killed a
couple of these processes. When I killed the 2nd all the remaining
disappeared. What does this indicate?
*** MSDTC is used for Distributed Transactions. If the IP address is not
from a machine
that you "Trust" , then this machine should be blocked.
When I have observed many connections on port 1433 the state is usually
WAIT_STATE or FIN_2 something.
*** TCP sessions have various states. FIN, TIME_WAIT are all valid states.
See:
137984 TCP Connection States and Netstat Output
http://support.microsoft.com/?id=137984
The server is a leased server at an ISP (Interland) that I administer
remotely via Terminal Services. There are 2 NICs: 1 to their LAN and 1
to the Internet.
*** Your provider should also have Firewall to prevent machines
from establishing connections. If they don't you should seriously consider
Publishing SQL from ISA Server.
For firewall I hoped to use IPSEC. I assume that 'Standard SQL
Security' means the same to you as to me (passwords, limit user
capabilities) and the answer is Yes.
*** If you only have certain valid clients that should be connecting to
your SQL Server
accross the internet, then IPSEc or a VPN could be used. Otherwise, since
your server
is open and using Standard Security, you'll be open to password guessing
attacks.
If the connection state is not "ESTABLISHED' does that mean that other
devices are attempting to connect unsuccessfully? Could these be port
scanner viruses or similar?
*** See TCP Connection States kb.
The msdtc.exe ESTABLISHED connections caused me more concern. Does that
mean my server and another device were 'handshaking' or exchanging
data?
*** If your SQL Server is not using Distributed Transactions, Stop the
MSDTC Server Service.
Thanks,
Kevin McDonnell
Microsoft Corporation
This posting is provided AS IS with no warranties, and confers no rights.|||Kevin,
Thanks for your help.
Since I stopped the MSDTC Server Service the spurious connections on Port 14
33 and 1025 have not reoccurred. I still have some on Port 21 that look like
viruses trying to find an FTP connection. I'll use IPSEC to filter those.
Mikesql

Needing to run reporting services on SQL Server 2000

Hi
I want to install SQL Server 2000 Enterprise Edition on Windows XP but it isn't installing.
I have installed SQL Server 2000 Personal Edition but Reporting Services are not installing on the server.
PLZ some one help me out.

SQL 2000 Enterprise Edition is not supported on XP. Please see this link for requirements:

http://www.microsoft.com/sql/prodinfo/previousversions/system-requirements.mspx

Thanks,
Sam Lester (MSFT)

Needed to connect Remote DB by Client Tool (Enterprise Manager)

Here is what I knew: usr/pwd, Domain, Server and Database.

How to set Enterprise Manager, then can I connect to DB ?

Take a look at the following:

Connecting to your SQL Database using Enterprise Manager

HTH

Needed software for designing reports

We currently have SQL 2005 (enterprise) reporting services hitting an
odbc based backend (mysql usually). It works great. I run Visual
Studio Pro (and have msdn sql server on my local machine), develop
reports locally and deploy them.
I know the express edition of SQL 2005 reporting services cannot use
an ODBC datasource. Only the standard and up versions.
My question is: For other employees to create and deploy reports what
software (and licenses) are needed? Ideally, I'd love to just deploy
Visual Studio Express and SQL 2005 Express for the users that need to
modify reports, but I have a feeling that won't work...
any ideas?
thank you!If they are just designing reports then you can give the BI Studio (not sure
if that is what it is called). When you install the designer it installs a
version of VS if you do not have it installed. This is free. Also, there are
no licensing issue either. RS is licensed by the server, not by where you
have the client tools installed.
You will want to have a server they can deploy to as part of the development
process, however, for most of their testing they can do this from the
development environment.
If the users can't handle Report Designer (which is really a developer
targeted tool) then check out report builder.
Note, you do not need SQL Server install locally in order to design reports.
Bruce Loehle-Conger
MVP SQL Server Reporting Services
<eric.gay@.gmail.com> wrote in message
news:1179174786.894457.142220@.h2g2000hsg.googlegroups.com...
> We currently have SQL 2005 (enterprise) reporting services hitting an
> odbc based backend (mysql usually). It works great. I run Visual
> Studio Pro (and have msdn sql server on my local machine), develop
> reports locally and deploy them.
> I know the express edition of SQL 2005 reporting services cannot use
> an ODBC datasource. Only the standard and up versions.
> My question is: For other employees to create and deploy reports what
> software (and licenses) are needed? Ideally, I'd love to just deploy
> Visual Studio Express and SQL 2005 Express for the users that need to
> modify reports, but I have a feeling that won't work...
> any ideas?
> thank you!
>

Need to update up to 3000 user DB

Hi,
We have an SQL 2005 Server with the whole enterprise data.
We are building an application that requires a per user database with its
own data (horizontal filtering).
Our first idea is to create 3000 mdf files, one db for each user. This DB is
created by copying a "model" db with the data structure, then using
SqlExpress's AttachDbFilename feature to connect to user db and fill it with
filtered data from the main server. Each user DB has the main server in its
linked servers in order to be able to query the main server from the user DB
and compare data from both sides (to find the differences). Subsequent
updates are done using the same SQL queries. MDF files are stored on a
server that has SQL Express installed to which the app connects.
This solution works well but the problem is that each user DB update (near
300 Sql requests for 50 mb results DB) requires at least 30 seconds to
execute. 3000 users means 25 hours to complete the whole process. The main
server stay at low CPU level, but the "express" server quickly increase its
CPU usage (near 100% when 4 updating threads are working simultaneously),
but we cannot buy a 64 cpu computer ;).
We are looking for a faster solution. Do you have any idea which could help
us either to optimize our solution or even another method ?
We do not want to use the Sql Server replication because it is not
applicable in our environment.
Buying other SQL Server is possible but must be justified.
Thanks,
SteveSteve
> Our first idea is to create 3000 mdf files, one db for each user. This DB
> is created by copying a "model" db with the data structure, then using
Actually you have not explained what are you trying to achive?
What is if the data will be stored in one database? Where do user locate?
How do they access the database?
We have many customers installed on their workstations MSDE/EXPRESS and just
pushed the data (as a publisher) from pur main server . Yep , it is a
replication which sometimes has some problems that hard to solve.
"Steve B." <steve_beauge@.com.msn_swap_com_and_msn> wrote in message
news:%23%23SBw8nTGHA.6048@.TK2MSFTNGP11.phx.gbl...
> Hi,
> We have an SQL 2005 Server with the whole enterprise data.
> We are building an application that requires a per user database with its
> own data (horizontal filtering).
> Our first idea is to create 3000 mdf files, one db for each user. This DB
> is created by copying a "model" db with the data structure, then using
> SqlExpress's AttachDbFilename feature to connect to user db and fill it
> with filtered data from the main server. Each user DB has the main server
> in its linked servers in order to be able to query the main server from
> the user DB and compare data from both sides (to find the differences).
> Subsequent updates are done using the same SQL queries. MDF files are
> stored on a server that has SQL Express installed to which the app
> connects.
> This solution works well but the problem is that each user DB update (near
> 300 Sql requests for 50 mb results DB) requires at least 30 seconds to
> execute. 3000 users means 25 hours to complete the whole process. The main
> server stay at low CPU level, but the "express" server quickly increase
> its CPU usage (near 100% when 4 updating threads are working
> simultaneously), but we cannot buy a 64 cpu computer ;).
> We are looking for a faster solution. Do you have any idea which could
> help us either to optimize our solution or even another method ?
> We do not want to use the Sql Server replication because it is not
> applicable in our environment.
> Buying other SQL Server is possible but must be justified.
> Thanks,
> Steve
>|||The Sql express db are used to store an image of data that each user
require.
The actual "end user" application is a mobile application in which a sql
mobile DB is synchronized against a http application that will make all the
required work to ensure the sql mobile DB has exactly the same data than the
Sql express one.
This "intermediary" sql express db is required since the MS Merge
replication between sql 2005 and mobile device can't be used because of
replication limitations.
Since the main DB is quite big, we want to compute differences for each user
within the lan, before sending it to the device (that's why we want to use
mdf files that are exact copy of the data for the user, which is easier to
synchronize since it is 1-1 with the device).
Thanks,
Steve
"Uri Dimant" <urid@.iscar.co.il> a écrit dans le message de news:
%23LSWwGoTGHA.4956@.TK2MSFTNGP09.phx.gbl...
> Steve
>> Our first idea is to create 3000 mdf files, one db for each user. This DB
>> is created by copying a "model" db with the data structure, then using
> Actually you have not explained what are you trying to achive?
> What is if the data will be stored in one database? Where do user locate?
> How do they access the database?
> We have many customers installed on their workstations MSDE/EXPRESS and
> just pushed the data (as a publisher) from pur main server . Yep , it is a
> replication which sometimes has some problems that hard to solve.
>
>
>
> "Steve B." <steve_beauge@.com.msn_swap_com_and_msn> wrote in message
> news:%23%23SBw8nTGHA.6048@.TK2MSFTNGP11.phx.gbl...
>> Hi,
>> We have an SQL 2005 Server with the whole enterprise data.
>> We are building an application that requires a per user database with its
>> own data (horizontal filtering).
>> Our first idea is to create 3000 mdf files, one db for each user. This DB
>> is created by copying a "model" db with the data structure, then using
>> SqlExpress's AttachDbFilename feature to connect to user db and fill it
>> with filtered data from the main server. Each user DB has the main server
>> in its linked servers in order to be able to query the main server from
>> the user DB and compare data from both sides (to find the differences).
>> Subsequent updates are done using the same SQL queries. MDF files are
>> stored on a server that has SQL Express installed to which the app
>> connects.
>> This solution works well but the problem is that each user DB update
>> (near 300 Sql requests for 50 mb results DB) requires at least 30 seconds
>> to execute. 3000 users means 25 hours to complete the whole process. The
>> main server stay at low CPU level, but the "express" server quickly
>> increase its CPU usage (near 100% when 4 updating threads are working
>> simultaneously), but we cannot buy a 64 cpu computer ;).
>> We are looking for a faster solution. Do you have any idea which could
>> help us either to optimize our solution or even another method ?
>> We do not want to use the Sql Server replication because it is not
>> applicable in our environment.
>> Buying other SQL Server is possible but must be justified.
>> Thanks,
>> Steve
>

Need to update up to 3000 user DB

Hi,
We have an SQL 2005 Server with the whole enterprise data.
We are building an application that requires a per user database with its
own data (horizontal filtering).
Our first idea is to create 3000 mdf files, one db for each user. This DB is
created by copying a "model" db with the data structure, then using
SqlExpress's AttachDbFilename feature to connect to user db and fill it with
filtered data from the main server. Each user DB has the main server in its
linked servers in order to be able to query the main server from the user DB
and compare data from both sides (to find the differences). Subsequent
updates are done using the same SQL queries. MDF files are stored on a
server that has SQL Express installed to which the app connects.
This solution works well but the problem is that each user DB update (near
300 Sql requests for 50 mb results DB) requires at least 30 seconds to
execute. 3000 users means 25 hours to complete the whole process. The main
server stay at low CPU level, but the "express" server quickly increase its
CPU usage (near 100% when 4 updating threads are working simultaneously),
but we cannot buy a 64 cpu computer ;).
We are looking for a faster solution. Do you have any idea which could help
us either to optimize our solution or even another method ?
We do not want to use the Sql Server replication because it is not
applicable in our environment.
Buying other SQL Server is possible but must be justified.
Thanks,
SteveSteve
> Our first idea is to create 3000 mdf files, one db for each user. This DB
> is created by copying a "model" db with the data structure, then using
Actually you have not explained what are you trying to achive?
What is if the data will be stored in one database? Where do user locate?
How do they access the database?
We have many customers installed on their workstations MSDE/EXPRESS and just
pushed the data (as a publisher) from pur main server . Yep , it is a
replication which sometimes has some problems that hard to solve.
"Steve B." <steve_beauge@.com.msn_swap_com_and_msn> wrote in message
news:%23%23SBw8nTGHA.6048@.TK2MSFTNGP11.phx.gbl...
> Hi,
> We have an SQL 2005 Server with the whole enterprise data.
> We are building an application that requires a per user database with its
> own data (horizontal filtering).
> Our first idea is to create 3000 mdf files, one db for each user. This DB
> is created by copying a "model" db with the data structure, then using
> SqlExpress's AttachDbFilename feature to connect to user db and fill it
> with filtered data from the main server. Each user DB has the main server
> in its linked servers in order to be able to query the main server from
> the user DB and compare data from both sides (to find the differences).
> Subsequent updates are done using the same SQL queries. MDF files are
> stored on a server that has SQL Express installed to which the app
> connects.
> This solution works well but the problem is that each user DB update (near
> 300 Sql requests for 50 mb results DB) requires at least 30 seconds to
> execute. 3000 users means 25 hours to complete the whole process. The main
> server stay at low CPU level, but the "express" server quickly increase
> its CPU usage (near 100% when 4 updating threads are working
> simultaneously), but we cannot buy a 64 cpu computer ;).
> We are looking for a faster solution. Do you have any idea which could
> help us either to optimize our solution or even another method ?
> We do not want to use the Sql Server replication because it is not
> applicable in our environment.
> Buying other SQL Server is possible but must be justified.
> Thanks,
> Steve
>|||The Sql express db are used to store an image of data that each user
require.
The actual "end user" application is a mobile application in which a sql
mobile DB is synchronized against a http application that will make all the
required work to ensure the sql mobile DB has exactly the same data than the
Sql express one.
This "intermediary" sql express db is required since the MS Merge
replication between sql 2005 and mobile device can't be used because of
replication limitations.
Since the main DB is quite big, we want to compute differences for each user
within the lan, before sending it to the device (that's why we want to use
mdf files that are exact copy of the data for the user, which is easier to
synchronize since it is 1-1 with the device).
Thanks,
Steve
"Uri Dimant" <urid@.iscar.co.il> a crit dans le message de news:
%23LSWwGoTGHA.4956@.TK2MSFTNGP09.phx.gbl...
> Steve
> Actually you have not explained what are you trying to achive?
> What is if the data will be stored in one database? Where do user locate?
> How do they access the database?
> We have many customers installed on their workstations MSDE/EXPRESS and
> just pushed the data (as a publisher) from pur main server . Yep , it is a
> replication which sometimes has some problems that hard to solve.
>
>
>
> "Steve B." <steve_beauge@.com.msn_swap_com_and_msn> wrote in message
> news:%23%23SBw8nTGHA.6048@.TK2MSFTNGP11.phx.gbl...
>

Need to update up to 3000 user DB

Hi,
We have an SQL 2005 Server with the whole enterprise data.
We are building an application that requires a per user database with its
own data (horizontal filtering).
Our first idea is to create 3000 mdf files, one db for each user. This DB is
created by copying a "model" db with the data structure, then using
SqlExpress's AttachDbFilename feature to connect to user db and fill it with
filtered data from the main server. Each user DB has the main server in its
linked servers in order to be able to query the main server from the user DB
and compare data from both sides (to find the differences). Subsequent
updates are done using the same SQL queries. MDF files are stored on a
server that has SQL Express installed to which the app connects.
This solution works well but the problem is that each user DB update (near
300 Sql requests for 50 mb results DB) requires at least 30 seconds to
execute. 3000 users means 25 hours to complete the whole process. The main
server stay at low CPU level, but the "express" server quickly increase its
CPU usage (near 100% when 4 updating threads are working simultaneously),
but we cannot buy a 64 cpu computer ;).
We are looking for a faster solution. Do you have any idea which could help
us either to optimize our solution or even another method ?
We do not want to use the Sql Server replication because it is not
applicable in our environment.
Buying other SQL Server is possible but must be justified.
Thanks,
Steve
Steve
> Our first idea is to create 3000 mdf files, one db for each user. This DB
> is created by copying a "model" db with the data structure, then using
Actually you have not explained what are you trying to achive?
What is if the data will be stored in one database? Where do user locate?
How do they access the database?
We have many customers installed on their workstations MSDE/EXPRESS and just
pushed the data (as a publisher) from pur main server . Yep , it is a
replication which sometimes has some problems that hard to solve.
"Steve B." <steve_beauge@.com.msn_swap_com_and_msn> wrote in message
news:%23%23SBw8nTGHA.6048@.TK2MSFTNGP11.phx.gbl...
> Hi,
> We have an SQL 2005 Server with the whole enterprise data.
> We are building an application that requires a per user database with its
> own data (horizontal filtering).
> Our first idea is to create 3000 mdf files, one db for each user. This DB
> is created by copying a "model" db with the data structure, then using
> SqlExpress's AttachDbFilename feature to connect to user db and fill it
> with filtered data from the main server. Each user DB has the main server
> in its linked servers in order to be able to query the main server from
> the user DB and compare data from both sides (to find the differences).
> Subsequent updates are done using the same SQL queries. MDF files are
> stored on a server that has SQL Express installed to which the app
> connects.
> This solution works well but the problem is that each user DB update (near
> 300 Sql requests for 50 mb results DB) requires at least 30 seconds to
> execute. 3000 users means 25 hours to complete the whole process. The main
> server stay at low CPU level, but the "express" server quickly increase
> its CPU usage (near 100% when 4 updating threads are working
> simultaneously), but we cannot buy a 64 cpu computer ;).
> We are looking for a faster solution. Do you have any idea which could
> help us either to optimize our solution or even another method ?
> We do not want to use the Sql Server replication because it is not
> applicable in our environment.
> Buying other SQL Server is possible but must be justified.
> Thanks,
> Steve
>
|||The Sql express db are used to store an image of data that each user
require.
The actual "end user" application is a mobile application in which a sql
mobile DB is synchronized against a http application that will make all the
required work to ensure the sql mobile DB has exactly the same data than the
Sql express one.
This "intermediary" sql express db is required since the MS Merge
replication between sql 2005 and mobile device can't be used because of
replication limitations.
Since the main DB is quite big, we want to compute differences for each user
within the lan, before sending it to the device (that's why we want to use
mdf files that are exact copy of the data for the user, which is easier to
synchronize since it is 1-1 with the device).
Thanks,
Steve
"Uri Dimant" <urid@.iscar.co.il> a crit dans le message de news:
%23LSWwGoTGHA.4956@.TK2MSFTNGP09.phx.gbl...
> Steve
> Actually you have not explained what are you trying to achive?
> What is if the data will be stored in one database? Where do user locate?
> How do they access the database?
> We have many customers installed on their workstations MSDE/EXPRESS and
> just pushed the data (as a publisher) from pur main server . Yep , it is a
> replication which sometimes has some problems that hard to solve.
>
>
>
> "Steve B." <steve_beauge@.com.msn_swap_com_and_msn> wrote in message
> news:%23%23SBw8nTGHA.6048@.TK2MSFTNGP11.phx.gbl...
>