Help Cannot resolve collation conflict for equal to operation.

Hello,
I am trying to setup replication for a database and I get the following
error. This is the content database of WSS. I noticed that the collation was
set for: Latin1_General_CI_AS_KS_WS which is diffent from our normal
collation setting SQL_Latin1_General_CP1_CI_AS, however I created the
rplDatabase with the same collation. Any help would be greatly appreciated.
Jake
Error Message:
The schema script '\\SERVER\D$\Program Files\Microsoft SQL
Server\MSSQL$DBSERVER\ReplData\unc\SERVER$DBSERVER _STS_WCPC-L-02_Content_STS
_WCPC_L_02_Content\20040416221336\proc_FindDocs_16 .sch' could not be
propagated to the subscriber.
Error Details:
The schema script '\\SERVER\D$\Program Files\Microsoft SQL
Server\MSSQL$DBSERVER\ReplData\unc\SERVER$DBSERVER _STS_WCPC-L-02_Content_STS
_WCPC_L_02_Content\20040416221336\proc_FindDocs_16 .sch' could not be
propagated to the subscriber.
(Source: Merge Replication Provider (Agent); Error number: -2147201001)
Cannot resolve collation conflict for equal to operation.
(Source: SERVER\DBSERVER (Data source); Error number: 446)
are you replicating to a SQL 7 database?
If you so you to configure your publication for this in the Specify
Subscriber Types dialog box of the create publication wizard. Otherwise in
the specify articles dialog of the same wizard click on the browse button to
the right of the articles and in the snpashot tab, select the collation tab.
"Jake" <rondican@.hotmail.com> wrote in message
news:%23RV1LCEJEHA.2380@.TK2MSFTNGP09.phx.gbl...
> Hello,
> I am trying to setup replication for a database and I get the
following
> error. This is the content database of WSS. I noticed that the collation
was
> set for: Latin1_General_CI_AS_KS_WS which is diffent from our normal
> collation setting SQL_Latin1_General_CP1_CI_AS, however I created the
> rplDatabase with the same collation. Any help would be greatly
appreciated.
> Jake
> Error Message:
> The schema script '\\SERVER\D$\Program Files\Microsoft SQL
>
Server\MSSQL$DBSERVER\ReplData\unc\SERVER$DBSERVER _STS_WCPC-L-02_Content_STS
> _WCPC_L_02_Content\20040416221336\proc_FindDocs_16 .sch' could not be
> propagated to the subscriber.
> Error Details:
> The schema script '\\SERVER\D$\Program Files\Microsoft SQL
>
Server\MSSQL$DBSERVER\ReplData\unc\SERVER$DBSERVER _STS_WCPC-L-02_Content_STS
> _WCPC_L_02_Content\20040416221336\proc_FindDocs_16 .sch' could not be
> propagated to the subscriber.
> (Source: Merge Replication Provider (Agent); Error number: -2147201001)
> ----
--
> --
> Cannot resolve collation conflict for equal to operation.
> (Source: SERVER\DBSERVER (Data source); Error number: 446)
> ----
--
> --
>
|||Jake,
Am I right to conclude that you have included the stored procedure
[proc_FindDocs.....] as an article in your publication? If yes, then the
error is occurring while applying the stored procedure to the subscriber,
because it contains code that is comparing two values (e.g. columns or
variables) of different data types. You could try to apply the stored
procedure script manually (through Query Analyser) to the subscriber
database to see if you get the same error. If yes, then analyse the stored
procedure to see which columns are involved and check their collation. This
should give you an indication of where the collation problem is occurring.
Hope the above helps.
Raj Moloye

Help after the server was renamed before replication was taken dow

Is there an easy way to take apart replication when a server has been
renamed. The sever in question was a subscriber to one publication and the
publisher of another. Is there an easy way to drop the relvant subscriptions
publications now that this server has been renamed. As I presume sql does not
just adapt to the new name?
I concurr with Paul. Replication is sensitive to the publisher name, so
renaming the publisher breaks the agents.
You are best to (as Paul suggests) script out and rebuild your publications
and subscriptions.
Hilary Cotter
Looking for a SQL Server replication book?
Now available for purchase at:
http://www.nwsu.com/0974973602.html
"Paul Ibison" <Paul.Ibison@.Pygmalion.Com> wrote in message
news:1d5101c4d7c0$4c921030$a501280a@.phx.gbl...[vbcol=seagreen]
>I would script out the replication settings then disable
> replication on the server. Make sure that SQL Server has
> been reconfigured:
> Use Master
> go
> Select @.@.Servername
> This should return your current server name but if it
> returns NULL then try:
> Use Master
> go
> Sp_DropServer 'OldName'
> GO
> Use Master
> go
> Sp_Addserver 'NewName', 'local'
> GO
> Stop and Start SQL Services
> Then edit your scripts accordingly with the new
> servername, reenable replication on the server and run
> the scripts.
> Rgds,
> Paul Ibison SQL Server MVP, www.replicationanswers.com
> (recommended sql server 2000 replication book:
> http://www.nwsu.com/0974973602p.html)
> server has been
> publication and the
> relvant subscriptions
> presume sql does not

HELP !Please . Urgent

after i done the replication from sql server 7.0(pub/dist) with sql server
2000(subscription) there will promt out the error like below :
Agents Last Action
Snapshot No such interface supported.NOTE:The step was retried
the
requested number of times(10)without succeeding.
The step failed.
Log Reader No such interface supported.NOTE:The step was retried
the
requested number of times(10)without succeeding.
The step failed.
202.168.xx:XXXXX No such interface supported.NOTE:The step was retried the
requested number of times(10)without succeeding.
The step failed.
ALL ERROR!!!!!
Please help...
thanks a lot
Sam,
are you doing merge replication by any chance - this configuration is not
supported.
Cheers,
Paul Ibison SQL Server MVP, www.replicationanswers.com .
|||It sounds like your com components did not get registered. Did you apply an
sp recently? If so you should reapply it.
If not, is this relevant?
http://groups.google.com/group/microsoft.public.sqlserver.replication/msg/6252b9649b296145
Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html
Looking for a FAQ on Indexing Services/SQL FTS
http://www.indexserverfaq.com
"sam" <sam@.discussions.microsoft.com> wrote in message
news:F09CBB95-050E-4704-AEB0-4C8B0AE77809@.microsoft.com...
> after i done the replication from sql server 7.0(pub/dist) with sql server
> 2000(subscription) there will promt out the error like below :
> Agents Last Action
> Snapshot No such interface supported.NOTE:The step was
> retried
> the
> requested number of times(10)without succeeding.
> The step failed.
> Log Reader No such interface supported.NOTE:The step was retried
> the
> requested number of times(10)without succeeding.
> The step failed.
> 202.168.xx:XXXXX No such interface supported.NOTE:The step was retried
> the
> requested number of times(10)without succeeding.
> The step failed.
> ALL ERROR!!!!!
> Please help...
> thanks a lot
>
|||I doing transactional replication.
Thanks for your replied.
"Paul Ibison" wrote:

> Sam,
> are you doing merge replication by any chance - this configuration is not
> supported.
> Cheers,
> Paul Ibison SQL Server MVP, www.replicationanswers.com .
>
>
|||OK - please see if this applies:
[url]http://groups.google.co.uk/group/microsoft.public.sqlserver.replication/browse_thread/thread/bdd1dfac50797c18/190e2af45971b1cb?lnk=st&q=No+such+interface+suppor ted+sql+server+replication&rnum=1&hl=en#190e2af459 71b1cb[/url]
Cheers,
Paul Ibison SQL Server MVP, www.replicationanswers.com .

Help - What name does the SQL Replication use?

Does anyone know if SQL 2000 Replication uses NetBios or Instance Name when
setting up a new replication partner? The reason I'm asking is because we've
got a server with a hostname of 16 characters long. The instance name is the
same as the hostname, which is also 16 characters long. When trying to setup
a replication, it fails unless we change the server name (and the instance
name) to 15 characters long. I thought SQL is using the Instance name, not
the Netbios name for replication? It seems that Netbios name is being used...
This is a very complex question. It uses the name that it was registered
with in Enterprise Manager. This should be the NetBIOS name of the machine +
the instance name. When it is mapping a drive or connecting to the FTP
server it uses the NetBIOS name, unless you specified your ftp server was a
FQDN.
Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html
Looking for a FAQ on Indexing Services/SQL FTS
http://www.indexserverfaq.com
"Monica" <Monica@.discussions.microsoft.com> wrote in message
news:60FAC952-AB91-4EBB-AF70-10F904A50087@.microsoft.com...
> Does anyone know if SQL 2000 Replication uses NetBios or Instance Name
> when
> setting up a new replication partner? The reason I'm asking is because
> we've
> got a server with a hostname of 16 characters long. The instance name is
> the
> same as the hostname, which is also 16 characters long. When trying to
> setup
> a replication, it fails unless we change the server name (and the instance
> name) to 15 characters long. I thought SQL is using the Instance name, not
> the Netbios name for replication? It seems that Netbios name is being
> used...
|||Hi Hilary,
Here's what we tried and it seems to fix the problem:
1. Re-register the 16 character long SQL server in the EM
2. Change the authentication to SQL/Windows. It was set to "Windows only".
3. Drop all the existing replication partners (those were setup with the
15-character SQL registration).
4. Re-created the replication partners.
Everything seems to work fine.
Do you know if SQL replication uses RPC remote connect to the "subscriber"?
(I can't find it being documented so far.) If so, then it makes sense.
Monica
"Hilary Cotter" wrote:

> This is a very complex question. It uses the name that it was registered
> with in Enterprise Manager. This should be the NetBIOS name of the machine +
> the instance name. When it is mapping a drive or connecting to the FTP
> server it uses the NetBIOS name, unless you specified your ftp server was a
> FQDN.
>
> --
> Hilary Cotter
> Looking for a SQL Server replication book?
> http://www.nwsu.com/0974973602.html
> Looking for a FAQ on Indexing Services/SQL FTS
> http://www.indexserverfaq.com
> "Monica" <Monica@.discussions.microsoft.com> wrote in message
> news:60FAC952-AB91-4EBB-AF70-10F904A50087@.microsoft.com...
>
>

Help - the replication Failed to Run

error i get from replication when i force them to run the process.
mircosoft sql - DMO(ODBC sql state:42000)
Error 14258: Cannot perform this operation while sqlserveragent is
starting.try again.
But i not restart any sqlserveragent And the sql server servies manager show
that all running.
hope can get asnwer as soon as posible .
thanks.
does this help?
http://support.microsoft.com/default.aspx/kb/303287
Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html
Looking for a FAQ on Indexing Services/SQL FTS
http://www.indexserverfaq.com
"sam" <sam@.discussions.microsoft.com> wrote in message
news:04A2D6ED-C57E-49EF-8952-50E082F5109F@.microsoft.com...
> error i get from replication when i force them to run the process.
> mircosoft sql - DMO(ODBC sql state:42000)
> Error 14258: Cannot perform this operation while sqlserveragent is
> starting.try again.
> But i not restart any sqlserveragent And the sql server servies manager
> show
> that all running.
> hope can get asnwer as soon as posible .
> thanks.

Help - SQL Server 2005->Compact Edition Replication

Hi All ...

Ihave been following the MS on-line tutorials to get replication setupbetween SQL Server 2005 and SQL Server Compact Edition. I get to thepoint where I run the Configure Web Synchronization and it is runningthrough the process where it is doing the automated configuration ofthe shared directory - assigning users etc.

I have attemptedthis step with a user account that I expressly created for snapshotsand with the Administrator account and get the following error message:

* The operation completed successfully. (Exception from HRESULT: 0x80070000)

Atfirst, I thought this was a rights issue due to the fact that I wasusing an account that I manually created. However, by using theAdministrator account (test only until I can figure this out) thatshould have removed any rights issues associated with accessing theshared directory where the snapshot will reside.

Any help in getting this figured out will be greatly appreciated.

David L. Collison
Any day above ground is a good day!Is the account local admin on the box? Also does the account belong to "sa" group?|||

Administrator is a local account - windows user.

"sa" is the database account - and yes, for test purposes I have the system using "sa". Again, tried a local account that I had created, but got error message. Once I can get working under Administrator and SA, I will then work on getting the local account working correctly.

|||

Whatever account you use should be BOTH local admin on the box as well as sysadmin for the db.

|||

I've made the changes to allow <localaccount>\Administrator to be a sysadmin on the DB. Still get the same error message.

Any other ideas?

|||Can you try re-initializing the snapshot? Also, when you login with the <Administrator> account and browse to the shared folder do you have write access to the folder?|||

As an update:

1. I have disabled Publishing and Distribution for the database and then reconfigured.

2. I have dropped the Snapshot Agent.

3. I have dropped and recreated the virtual folder off the Default Web Site.

4. I have switched between the Administrator account and a specific account I created to manage activity on the snapshots.

5. I have switched between the DB SA account and a specific DB account that I created to manage activity on the snapshots.

6. I have modified the rights of the DOMAIN Users Administrator and the specific account I created to have "Full Control" rights to the snapshot directory and to the virtual directory.

7. I have modified the DB rights of the DB users to have db_owner, db_datareader, db_datawriter.I continue to get the error:The operation completed successfully. (Exception from HRESULT: 0x80070000)(mscorlib)

8. I have attempted to create the virtual directory manually within the Default Web location and assign the rights and then run the wizard to no avail, I get the same error message.

I'm at my wits end on this.

|||

Let me explain. Create one domain account. Add the account to the Administrator group for the box. Also, add the same account to both the SQL Server's (publisher and subscriber) and give it "sa" rights (on both servers). Replication agents copy the files to the local folder and then copy it onto the subscriber. So the account that does all this needs to be local admin as well as "sa" on the SQL box. Grant the rights at the server level. Individual db_owner, db_reader is not required if you give the account "sa" rights at the server level. Also, make sure you are logged into the box where you are setting up replication. Just logging into the SQL server through enterprise manager and setting up replication will not work. You need to be logged onto the server locally. When you are done, check each of the replication jobs. See what account each job is running under. It should be the same domain account that was created earlier.

|||

Dinakar

I've setup the accounts as you have indicated and I'm still getting the error message on the last page of the wizard. I don't want to, but I'm thinking I may have to blow away SQL Server and reinstall everything - which causes me a great deal of heartburn because of all of the development activity I have on this server, not to mention having to reconfigure SQL Server Reporting Services.

Any further ideas before I take this fatalistic step?

Thanks

|||

Folks ...

I finally ended up initiating an incident with Microsoft on this issue and have found a resolution to the problem.

1) You must create a Network share - I was forced to use the NET SHARE command from a Command Prompt - for whatever reason, the system would not recognize the share created via Windows.

2) Next, during the Wizard, you get to the screen labeld "Snapshot Share Access". In this window, I had browsed to the physical directory directory where the snapshot exists - THIS IS NOT VALID (note to MS - remove the Browse function). Here, you must enter the share created in the previous step -\\ComputerName\ShareName

The wizard then completes normally.

Side note, during all of this, MS tech support had me delete the Snapshot and recreate the snapshot. Once I deleted the snapshot and was attempting to recreate the snapshot, I began receiving errors that certain system tables did not exist. I was forced to manually remove the replication functionality by opening SQL Server Management Studio and executing the following:

1) USE <dbName>

2) EXEC sp_removedbreplication '<dbName>'

I was then able to recreate the snapshot.

Hope this helps others out who are having similar issues.

HELP - Replication Error

We have about 52 clients on SQL Server 2005 using anonymous pull
subscriptions (2 subs each).
One of these clients had an issue with disk space so we had to recreate the
subscriptions for the 2 databases. One subscription went fine but the second
is getting the following error:
The specified subscription type is invalid. Verify that the
-SubscriptionType parameter for the Merge Agent has been correctly specified.
(Source: MSSQL_REPL, Error number: MSSQL_REPL-2147201020
The -SubscriptionType parameter is set to 2. Nothing is different from the
other 100+ subscriptions we have created. I compared the working
subscription's step to the failing sub and they are identical except for the
db and pub name.
I have been unsuccesful in finding anything on the Internet. Does anybody
have a solution?
Thanks in advance!
can you try a subscriptiontype of 1?
Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html
Looking for a FAQ on Indexing Services/SQL FTS
http://www.indexserverfaq.com
"RichardD" <RichardD@.discussions.microsoft.com> wrote in message
news:6733A022-8948-4D15-A116-58A5C81CCA94@.microsoft.com...
> We have about 52 clients on SQL Server 2005 using anonymous pull
> subscriptions (2 subs each).
> One of these clients had an issue with disk space so we had to recreate
> the
> subscriptions for the 2 databases. One subscription went fine but the
> second
> is getting the following error:
> The specified subscription type is invalid. Verify that the
> -SubscriptionType parameter for the Merge Agent has been correctly
> specified.
> (Source: MSSQL_REPL, Error number: MSSQL_REPL-2147201020
> The -SubscriptionType parameter is set to 2. Nothing is different from the
> other 100+ subscriptions we have created. I compared the working
> subscription's step to the failing sub and they are identical except for
> the
> db and pub name.
> I have been unsuccesful in finding anything on the Internet. Does anybody
> have a solution?
> Thanks in advance!

Help - Replication Architecture Suggestions

Hi,
I need some help in designing the physical architecture of my database
system for replication. Here is my scenario
I have 1 server each in North America, Europe and Asia. They will be
connected 99 % of the time. All the data needs to be replicated between these
servers. There is very less data which will create contain conflicts while
replicating.
What type of replication should i be having ?
Should i go with the local distributors ?
If there is a conflict and i have local distributors then how can that be
resolved between 3 subscribers ? If possible please provide examples
The above is one scenario
Now each of these servers will be connected by local servers running MSDE.
They do not have a guaranteed connection. They need to upload data and also
recieve fresh updates from these regional servers. There can be a conflict
between the data of 2 local servers which will need to be resolved before
uploading to the regional server. There will be specific conflict policies
defined.
How do i achieve the above ?
I would appreciate if someone can help me in this with examples. I have not
implemented replication till date.
Thanks in advance
Afaq
Afaq,
for this type of scenario, merge replication functions best. There would be
no need to republish and have ultiple distributors - just designate the
'main' server as the publisher (and distributor) and the others as
subscribers. From your description, conflicts can arise as the data is not
partitioned - merge replication will allow you to have column-level or
row-level conflict detection. By default, a change on the publisher will
'win' against a change on the subscribers, and if 2 subscribers conflict,
the first one to synchronize 'wins'. There are other algorithms that can be
used, and you can create your own conflict resolution routine if necessary.
For more info, have a look in BOL for "conflict resolution, overview".
HTH,
Paul Ibison
|||Paul,
Thanks for the reply. Another question i had was regarding the PK for the
tables. I wanted to have identities as the PK. After reading the posts i
realised that identities are not so easy. If i have GUID's then my indexing
performance goes down which is a issue for me.
Can someone suggest me how best i can use identities with merge replication.
I cn use the Not for replciation clause but for that i will have to specify a
range of identity values for a table. That is also acceptable. But how do i
manage when i reach the end of the range? Do i have to manually increase the
range? how to handle this when i have 500 msde's working together. how to
achieve the range on each of these servers.
Any help would be appreciated
Afaq
"Paul Ibison" wrote:

> Afaq,
> for this type of scenario, merge replication functions best. There would be
> no need to republish and have ultiple distributors - just designate the
> 'main' server as the publisher (and distributor) and the others as
> subscribers. From your description, conflicts can arise as the data is not
> partitioned - merge replication will allow you to have column-level or
> row-level conflict detection. By default, a change on the publisher will
> 'win' against a change on the subscribers, and if 2 subscribers conflict,
> the first one to synchronize 'wins'. There are other algorithms that can be
> used, and you can create your own conflict resolution routine if necessary.
> For more info, have a look in BOL for "conflict resolution, overview".
> HTH,
> Paul Ibison
>
>
|||Afaq,
500 MSDE's is a lot
As far as I can see, you have 3 choices:
(a) you can partition the PK values. In this way, each subscriber has an
identity column combined with another column as a PK. The identity column
functions as a normal one - no partitioning - and the other column
identifies the subscriber. In this way there will be no conflicts and no
need for ranges.
(b) you can use automatic range management. An integer column goes to
+2billion (ish), so you should be able to assign large ranges which will
never need to be updated. If you use small ranges, when they are full up the
check constraint will error and new records can't be entered until the
subscriber synchronizes and you want to avoid this possibility.
(c) you can manually assign the ranges. For this you'd need to set up each
subscriber separately and this might turn out to be too much work...
HTH,
Paul Ibison
|||I guess the first option is doable.
Now there is another issue
All my servers are publishers and subscribers. So please consider the
following example
Table A is an article in a publication on Server X. Server Y is a subscriber
to this pubication. Server Y also can be a publisher for Table A and Server X
can be a subscriber for the same. WHen i have a subscription on table A from
server Y i cant have a publication for the same table from Server Y.
How do i achieve the same? I need Table A to be a part of different
publications from different servers as they are in different time zones and
they will publish the data on different times. And we want each of the
servers to publish the data for the same table at different times.
I hope the issue is properly conveyed. Forgive me for asking this question
as i am a newbie in replication.
Thanks
Afaq
"Paul Ibison" wrote:

> Afaq,
> 500 MSDE's is a lot
> As far as I can see, you have 3 choices:
> (a) you can partition the PK values. In this way, each subscriber has an
> identity column combined with another column as a PK. The identity column
> functions as a normal one - no partitioning - and the other column
> identifies the subscriber. In this way there will be no conflicts and no
> need for ranges.
> (b) you can use automatic range management. An integer column goes to
> +2billion (ish), so you should be able to assign large ranges which will
> never need to be updated. If you use small ranges, when they are full up the
> check constraint will error and new records can't be entered until the
> subscriber synchronizes and you want to avoid this possibility.
> (c) you can manually assign the ranges. For this you'd need to set up each
> subscriber separately and this might turn out to be too much work...
> HTH,
> Paul Ibison
>
>

HELP - Problem with Replication

Hello,

I have transactional replication running between two SQL Servers. It's a push subscription and the distributor is on the local publisher.

Recently after running for a year the replication stopped working and we get the following two errors.

1 - The process could not bulk copy into the table "History_Rep"

String Data, Right Truncation. Error Number 22001.

-

2 - The process could not bulk copy into the table "History_Rep"

Unexpected EOF encountered in BCP Data-File. Error Number S1000

The data types between the source and target tables are identical. I've tried emptying my source tables so that BCP doesn't have to send anything to the subscriber but it still fails. Do I maybe need to purge my distribution database and BCP log files? How do I go about doing that.

Thanks in advance Guys,

Hugo

Hi Hugo,

Once you have emptied the source table, you would need to reinitialize your subscription and re-generate the snapshot in order to 'remove' the bad bcp data from the system. Having said that, I can't help but suspect that there is a schema (column size) mismatch between the publisher and the subscriber based on the error message that you posted. Is this on SQL2000 (my guess) or SQL 2005?

-Raymond

|||

Raymond.

Yes, SQL Server 2000.

I've checked the column names, data types, sizes, three or four times - they're identical. How do I reinitialize the subscription - do I have to remove it and re-add it? Can it be done easily with Enterprise manager?

Thanks,

Hugo

|||

Hi Hugo,

You should be able to just right-click on your subscription node in Enterprise Manager and choose reinitialize subscription on the context manual. Does the problematic table have multiple text\image columns at the end by any chance?

-Raymond

|||

Hey Raymond,

When I right click on the subscription node, there's no option for re-initializing the subscription. I can't find that anywhere.

The table is all varchars and ints. No Text fields. There is a varchar 4000 second from the end where we put some XML data but that's it.

Thanks,

Hugo

|||

Hi Hugo,

My thinking was probably mixed up by how the SQL2005 UI is laid out, here is what worked for me using the SQL2000 Enterprise Manager:

For a well-known push or pull subscription:

1) Publisher Server->Expand Replication Folder->Expand Publications Folder->Select publication whose subscription(s) you want to reinitialize, all your subscriptions should show up on the right plane->now right-click on the subscription you want to reinitialize on the right-plane and you should be able to choose "Reinitialize" from the context menu.

For an anonymous pull subscription:

2) Subscriber Server->Expand Replication Folder->Select Subscriptions Folder, all your pull subscriptions should show up on the right-plane-> Right-click on the subscription you want to reinitialize on the right-plane and you should be able to choose "Reinitialize" from the context menu.

Hope that helps,

-Raymond

|||

Raymond,

Awesome - that worked. Not it says it's pending - why? How can I find out why it's pending - there's nothing in any logs that I can see.

Thanks for all your help,

Hugo

|||

Hey Raymond,

I also started the snapshot agent and it ran okay. But, my subscription still says pending and when I try to start the distribution agent I get the same error.

Hugo

|||

Have you restarted the snapshot agent?

-Raymond

|||

Sorry for the hasty reply, were there any rows bulk copied out by the snapshot agent for the problematic table?

-Raymond

|||

No need for apologies

Well, it said four articles were generated - I'll investigate more.

Hugo

|||

Hugo, something is not adding up correctly in your case (assuming that you have indeed deleted all data of the problematic table at the publisher), you can drop and re-add the subscription to make sure that it is not picking up a stale snapshot (I am out of ideas and my mind if that doesn't work). I am also wondering if you have different code pages between the publisher and the subscriber (or if you are using character mode snapshot) which would have caused the original bcp problem to happen. In any case, it would appear that the problem becomes extremely difficult to debug through forum postings, and so I would really appreciate if you can:

1) Open a support case with product support so they can gather sufficient information for us to investigation the issue further, or

2) Log a bug @. http://connect.microsoft.com/registration/registration.aspx with enough information (schema & data of problematic article) so we can investigate the issue properly.

Thanks much!

-Raymond

|||

Raymond,

We solved the problem last night - thanks. There was a program pumping data into the problematic table and once we shut that program down and followed your advice it worked. I think that maybe I have character mode of the snapshot going on because that has never worked. Can I change that - how?

Thanks again,

Hugo