Hi.
I still have two questions :)
First question :
Is the argument -ForceConvergenceLevel specific to SQL Server 2000?
Is there something special I got to install or configure to use this
argument...?
The argument does no seem to exist... Here are the arguments I can use :
(Sorry for the long list....)
replmerg [-?]
-Publisher server_name[\instance_name]
-PublisherDB publisher_database
-Publication publication
-Subscriber server_name[\instance_name]
-SubscriberDB subscriber_database
[-Continuous]
[-DefinitionFile def_path_and_file_name]
[-Distributor server_name[\instance_name]]
[-DistributorAddress distributor_address]
[-DistributorLogin distributor_login]
[-DistributorNetwork distributor_network]
[-DistributorPassword distributor_password]
[-DistributorSecurityMode [0|1]]
[-DownloadGenerationsPerBatch download_generations_per_batch]
[-DownloadReadChangesPerBatch download_read_changes_per_batch]
[-DownloadWriteChangesPerBatch download_write_changes_per_batch]
[-ExchangeType [1|2|3]]
[-FastRowCount [0|1]]
[-FileTransferType [0|1]]
[-FtpAddress ftp_address]
[-FtpPassword ftp_password]
[-FtpPort ftp_port]
[-FtpUserName ftp_user_name]
[-HistoryVerboseLevel [1|2|3]]
[-Hostname host_name]
[-LoginTimeOut login_time_out_seconds]
[-Output]
[-OutputVerboseLevel [0|1|2]]
[-PollingInterval polling_interval]
[-ProfileName profile_name]
[-PublisherAddress publisher_address]
[-PublisherLogin publisher_login]
[-PublisherNetwork publisher_network]
[-PublisherPassword publisher_password]
[-PublisherSecurityMode [0|1]]
[-QueryTimeOut query_time_out_seconds]
[-SubscriberDatabasePath subscriber_path]
[-SubscriberDBAddOption [0|1|2|3]]
[-SubscriberLogin subscriber_login]
[-SubscriberPassword subscriber_password
[-SubscriberSecurityMode [0|1]]
[-SubscriberType [0|1|2|3|4|5|6|7|8]]
[-SubscriptionType [0|1|2]]
[-UploadGenerationsPerBatch upload_generations_per_batch]
[-UploadReadChangesPerBatch upload_read_changes_per_batch]
[-UploadWriteChangesPerBatch upload_write_changes_per_batch]
[-Validate [0|1|2|3]]
[-ValidateInterval validate_interval]
I need your help to find this argument.
Second question :
I also have another problem. Maybe you could help me.
The problem is in the same merge replication.
Here are the steps :
1. I created a snapshot on the publisher.
2. I inserted data on the publisher.
3. The subscriber merged the data (with a dynamic filter). Only some rows of
the data inserted were merged (because of the filter).
4. I changed data at the publisher. I modified rows that caused rows that
were not merged to be merged.
5. I started the merge again and some rows that should be replicated are
not.
Some of the rows inserted in step 2 should now be filtered and merged.
It seems that the rows that were not transfered during a merge will never be
transfered even if some conditions in the filters suppose that they should
be transfered.
Example :
Tables :
T_Students : Table containing students of a language school.
T_Tutor : Table containing tutors.
T_Susbstitutes : Table that's used to say "this tutor replaces this other
tutor".
When a distant tutor merges with the publisher database, the merge transfers
his students... However, when the tutor replaces another tutor (when he his
sick), we want to transfer the students that belong to the other tutor. It
does not work properly, because the T_Student table has already been
filtered.
However, if this is the first merge and the tutor replaces another tutor,
the data is transfered.
The problem seems to be that we can not retrieve data changes made before
the last merge, even if the filter tells that they should be merged now.
I could reinitialize the subscription before each merge... but it's not the
perfect way to do this....
Is there another way to avoid this problem...?
Thanks a lot for your precious help.
Hugues
> YOu can add this to the "run agent" job step of the merge job.
> If you goto command prompt and type replmerg.exe -? you will see the help.
> > Thanks again for your answer.
> > However, I'd like to ask you another question.
> > Where can I find this parameter in Enterprise Manager or anywhere else?
> > I looked everywhere, even in the books online and on MSDN.
> > I can't find anything. Please help me :)
> > Thanks.
> > Hugues
> > > There is also another way to handle this. The merge agent has an
> > > argument -ForceConvergenceLevel
> > > If you use 1 it will evaluate all generations for you.
> > > > Hi.
> > > > I checked the numbers and the number at the subscriber was higher
than
> > the
> > > > number aat the publisher.
> > > > To correct the problem I just have to change this number to a lower
> > > number.
> > > > Your answer saved me a lot of time, I was getting desperate.
> > > > So, thanks for your help :)
> > > > Hugues
> > > > > check to see if the rowguid in MSMerge_contents on the source of
the
> > > > change
> > > > > belongs to a generation which is greater than the recgen at the
> > > > destination.
> > > > > You could find out the recgen value from MSmerge_replinfo on the
> > > > subscriber.
> > > > > > Hi everybody.
> > > > > > I'm having a strange problem.
> > > > > > My configuration is :
> > > > > > SQL Server 7 SP3
> > > > > > Publisher NT4, Subscriber Win98
> > > > > > Merge replication with dynamic filters
> > > > > > Pull (anonymous) subscription
> > > > > > The subscriber connects to the publisher on the Internet via a
> VPN.
> > > > > > I can merge any changes, so my setup is ok.
> > > > > > I can create a snapshot and insert data and the data is
correctly
> > > > merged.
> > > > > > HOWEVER, if I create the snapshot, insert data and wait for the
> next
> > > > > > day to do the merge, it says that no data needed to be merged.
> > > > > > No snapshot is scheduled to start for example a night. I have
> > verified
> > > > > > this.
> > > > > > I don't know what's happenning.
> > > > > > Does anyone know how I could correct this problem?
> > > > > > Thanks in advance for your help.
> > > > > > Hugues