Merge Replication!

Can someone please advise what would be appropriate method when you want to
implement the merge replication (Kind of a Real time) between different
databases sitting on separate servers.
Is Auto Incremental Field which is PK would be a problem when two users try
to enter records at same time. For example in table#1 user "A" enter some
values at row#3 and got the Id 3,
While at the same time the users "B" in table#2 enter some values at
row#3...would he got Id 3 ? and how it would get replicated.
Do I need to use GUID as PK or some other suggestion?
Shan
Merge replication does not provide real time replication, use automatic
identity range management to avoid the pk conflicts. To do this, drop all
publications where the table you want to bring under automatic identity
range management is publisher. In the publication wizard when you come to
the specify articles dialog box, click on the browse button to the right of
your table name, and select the automatic identity range management tab.
Select a publisher and subscriber range which matches that maximum number of
inserts which might occur between agent syncs, and then make sure your
agents do not run continuously.
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
"Shan" <Shan@.discussions.microsoft.com> wrote in message
news:C97F8918-2523-4064-A6EB-F7C1D7D67BE2@.microsoft.com...
> Can someone please advise what would be appropriate method when you want
to
> implement the merge replication (Kind of a Real time) between different
> databases sitting on separate servers.
> Is Auto Incremental Field which is PK would be a problem when two users
try
> to enter records at same time. For example in table#1 user "A" enter some
> values at row#3 and got the Id 3,
> While at the same time the users "B" in table#2 enter some values at
> row#3...would he got Id 3 ? and how it would get replicated.
> Do I need to use GUID as PK or some other suggestion?
> --
> Shan

Merge Replication With SQL Server Express (Getting Started)

Hi

This is my first time working with replication and I was wondering if someone could point me in the right direction. I am looking for some sample code or a Web page that describes how to do merge replication using SQL Server Express on the client machines and SQL Server 2005 on the backend. If someone could help, I would appreciate it. Thanks.

If you are familiar with replication on other SKUs (such as standard edition), setting up on Express is not that complicated. You can treat Express just as another SKU, with the following exceptions:

1) Express can only be a subscriber.

2) There is no SQL Server Agent on Express. Thus if you have pull subscription, you can use one of the following ways to initiate synchronization on Express:

a) Using Sync Manager

b) Using ActiveX

c) Calling replmerg.exe (for merge subscription) or distrib.exe (for transactional subscription) with the proper parameters directly.

Merge replication re-initialization

I have 2 questions,
1. In what circumstances a subscriber need a re-initialization?
2. When I re initialize the second time, Does it re applies the whole
snapshot to the subscriber again or applies only the delta corrections?
1) when you make a change to the publication, when a subscriber expires, or
when the publisher and subscriber have a data validation problem.
2) the entire snapshot - there is an option to upload changes which have
occurred on a subscriber since the last sync.
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
"Ravi Lobo" <RaviLobo@.discussions.microsoft.com> wrote in message
news:53A25921-8853-4BFB-B611-E41374D9390F@.microsoft.com...
>I have 2 questions,
> 1. In what circumstances a subscriber need a re-initialization?
> 2. When I re initialize the second time, Does it re applies the whole
> snapshot to the subscriber again or applies only the delta corrections?
>
|||Thank you Hilary!

Merge Replication Problems

I am having a heck of a time trying to get merge replication to work. I have two tables that consistently fail validation - even just rowcount after applying a snapshot that I created seconds before and at a time when there is no activity on the tables.

Is there a guide/info somewhere about how to troulbeshoot this? The tables are fairly straightforward (but 35,000 rows) and there are no filters set up. The remote database is connected over the internet and I was hoping to be able to get the two databases to synch continuously.

Also, when I create the merge agent, the wizard doesn't show the schedule option (i.e. whether to run continuously or on a schedule) and therefore I have to set the -continuous option on the merge agent manually.

I have been struggling with this for 2 weeks so any help/pointers would be most appreciated.

After more playing around, it appears that if I create a merge publication to a database that already has a continuous transactional subscription, I don't get the option to set the merge subscription to update the merge subscription continuously. I've now put all my tables into one big merge publication.... hopefully that will work.

Am I on the right track: in sql2000 is there some limitation that doesn't allow a continuous merge and a continuous transactional publication/subscription to and from the same publisher/subscriber database pair?

thanks....

|||

I don't know if this is the answer to my own question but: it appears that trying to get continuous merge and continuous transactional replication to coexist from the same publisher database to the same subscription database was my problem.

I have no idea why, nor whether this is documented somewhere, but now I have made all of my publications continuous merge (even though transactional would seem to be the reight choice for most of the tables) gets over all the problems I was having with the merge publications continually failing.

Can anyone confirm this? Or did I unknowingly stumble upon some other fix while reconfiguring all replication from scratch?

|||

if you want to do validation, you need to do it when activity on the subscriber and publisher are minimal, otherwise if changes are continuously happening on one end while validation is occuring, there's a chance you'll never be in sync.

If you stop the transactional replication, and stop all other changes, I bet your validation will pass.

Merge Replication Problems

I am having a heck of a time trying to get merge replication to work. I have two tables that consistently fail validation - even just rowcount after applying a snapshot that I created seconds before and at a time when there is no activity on the tables.

Is there a guide/info somewhere about how to troulbeshoot this? The tables are fairly straightforward (but 35,000 rows) and there are no filters set up. The remote database is connected over the internet and I was hoping to be able to get the two databases to synch continuously.

Also, when I create the merge agent, the wizard doesn't show the schedule option (i.e. whether to run continuously or on a schedule) and therefore I have to set the -continuous option on the merge agent manually.

I have been struggling with this for 2 weeks so any help/pointers would be most appreciated.

After more playing around, it appears that if I create a merge publication to a database that already has a continuous transactional subscription, I don't get the option to set the merge subscription to update the merge subscription continuously. I've now put all my tables into one big merge publication.... hopefully that will work.

Am I on the right track: in sql2000 is there some limitation that doesn't allow a continuous merge and a continuous transactional publication/subscription to and from the same publisher/subscriber database pair?

thanks....

|||

I don't know if this is the answer to my own question but: it appears that trying to get continuous merge and continuous transactional replication to coexist from the same publisher database to the same subscription database was my problem.

I have no idea why, nor whether this is documented somewhere, but now I have made all of my publications continuous merge (even though transactional would seem to be the reight choice for most of the tables) gets over all the problems I was having with the merge publications continually failing.

Can anyone confirm this? Or did I unknowingly stumble upon some other fix while reconfiguring all replication from scratch?

|||

if you want to do validation, you need to do it when activity on the subscriber and publisher are minimal, otherwise if changes are continuously happening on one end while validation is occuring, there's a chance you'll never be in sync.

If you stop the transactional replication, and stop all other changes, I bet your validation will pass.

merge replication only need passwrd first time

Hello
I have an applicatiuon on a pda where i can synchronize the pda with
sql server 2000
When I start the application on open the synchronize window i have to
enter a user id and password.
repl.InternetLogin = userId
repl.InternetPassword = Password
I start the synchronization and it works.
The second time i synchronize it still works even with no password.
Why is that? I have to stop the application and start it again for it
to work.
If I enetr a wrong password when I synchronize the first time, I gett e
message saying: Authentication failed on the computer running IIS
I would like to get this message all the time when I synchronize with
an incorrect password.
What can I be doing wrong.
What form of authentication are you using on your IIS Server. Is it possible
that you have a blank password for the account userID?
Hilary Cotter
Director of Text Mining and Database Strategy
RelevantNOISE.Com - Dedicated to mining blogs for business intelligence.
This posting is my own and doesn't necessarily represent RelevantNoise's
positions, strategies or opinions.
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
<jensenove@.netscape.net> wrote in message
news:1144351127.584971.98050@.t31g2000cwb.googlegro ups.com...
> Hello
> I have an applicatiuon on a pda where i can synchronize the pda with
> sql server 2000
> When I start the application on open the synchronize window i have to
> enter a user id and password.
> repl.InternetLogin = userId
> repl.InternetPassword = Password
>
> I start the synchronization and it works.
> The second time i synchronize it still works even with no password.
> Why is that? I have to stop the application and start it again for it
> to work.
> If I enetr a wrong password when I synchronize the first time, I gett e
> message saying: Authentication failed on the computer running IIS
> I would like to get this message all the time when I synchronize with
> an incorrect password.
> What can I be doing wrong.
>

Merge Replication increases the bulk insert time dramatically

You have triggers firing which populate the
MSmerge_contents table. You also have the merge agent
which tries to read this table. So, there are 2
additional features. You can't remove the triggers -
they're mandatory so try disabling the merge agent while
the bulk upload takes place. If this still takes a while,
try the same bulk load on an identical copy of the table
without the triggers - just to confirm that it is the
triggers and not the higher quantity of data that you now
have in the table, as placing rows according to the PK
value will by necessity take longer.
Rgds,
Paul Ibison SQL Server MVP, www.replicationanswers.com
(recommended sql server 2000 replication book:
http://www.nwsu.com/0974973602p.html)
I checked, and we didn't have FIRE_TRIGGERS in our bulk insert commands. The
data was still being replicated, however.
Can you programatically disable the merge agent and then re-enable it during
bulk inserts? If so, can you point me to the right commands?
When we turn off replication, the insertion times remain around 3 seconds.
"Paul Ibison" wrote:

> You have triggers firing which populate the
> MSmerge_contents table. You also have the merge agent
> which tries to read this table. So, there are 2
> additional features. You can't remove the triggers -
> they're mandatory so try disabling the merge agent while
> the bulk upload takes place. If this still takes a while,
> try the same bulk load on an identical copy of the table
> without the triggers - just to confirm that it is the
> triggers and not the higher quantity of data that you now
> have in the table, as placing rows according to the PK
> value will by necessity take longer.
> Rgds,
> Paul Ibison SQL Server MVP, www.replicationanswers.com
> (recommended sql server 2000 replication book:
> http://www.nwsu.com/0974973602p.html)
>
>
|||There's something I don't understand here. When you did
the bulk insert without firing the triggers, how is it
possible that the data could be merged? Perhaps when you
say the data was still being replicated, you mean that
the merge agent was running? Anyway, What are the times:
(a) no replication
(b) replication and no firing of triggers and the merge
agent on a schedule which means it is not currently
running
(c) replication and firing of triggers and the merge
agent on a schedule which means it is not currently
running
(d) replication and firing of triggers and the merge
agent currently running
Rgds,
Rgds,
Paul Ibison SQL Server MVP, www.replicationanswers.com
(recommended sql server 2000 replication book:
http://www.nwsu.com/0974973602p.html)
|||Hi Paul.
I do not understand either. However, the data is on both servers.
Here are some times:
a)No replication: average 1.5 seconds daily
d)Continuous Replication: 3,4,14,29,118 (averages of daily values over 5 days)
I will have to test the other two options.
I understand that adding RGUIDs will increase the time it takes to insert
data, but is an exponential increase expected?
"Paul Ibison" wrote:

> There's something I don't understand here. When you did
> the bulk insert without firing the triggers, how is it
> possible that the data could be merged? Perhaps when you
> say the data was still being replicated, you mean that
> the merge agent was running? Anyway, What are the times:
> (a) no replication
> (b) replication and no firing of triggers and the merge
> agent on a schedule which means it is not currently
> running
> (c) replication and firing of triggers and the merge
> agent on a schedule which means it is not currently
> running
> (d) replication and firing of triggers and the merge
> agent currently running
> Rgds,
> Rgds,
> Paul Ibison SQL Server MVP, www.replicationanswers.com
> (recommended sql server 2000 replication book:
> http://www.nwsu.com/0974973602p.html)
>
|||Hi Paul,
It turns out our MSmerge_contents table was over 2.5 million records.
And it's not automatically clearing up. We are inserting data (not using
bulk insert) every five minutes, so this is a critical concern for us.
Could this be the problem? How can we tell it to automatically clean up
after data is replicated?
"Paul Ibison" wrote:

> There's something I don't understand here. When you did
> the bulk insert without firing the triggers, how is it
> possible that the data could be merged? Perhaps when you
> say the data was still being replicated, you mean that
> the merge agent was running? Anyway, What are the times:
> (a) no replication
> (b) replication and no firing of triggers and the merge
> agent on a schedule which means it is not currently
> running
> (c) replication and firing of triggers and the merge
> agent on a schedule which means it is not currently
> running
> (d) replication and firing of triggers and the merge
> agent currently running
> Rgds,
> Rgds,
> Paul Ibison SQL Server MVP, www.replicationanswers.com
> (recommended sql server 2000 replication book:
> http://www.nwsu.com/0974973602p.html)
>

Merge Replication IIS Worker Process Error

Everyday between 18:00 and 20:00 nearly 1000 PDA Subsriber anonymously
synchronise via Merge Replication and at least two time he have the error :

IIS Worker Process
Faulting application w3wp.exe, version 6.0.3790.1830,
faulting module sscerp20.dll, version 2.0.7331.0,
fault address 0x000110f4.

And subscriber which synchronising meanwhile becomes suspect.

Can someone offer a suggestion as to the cause of and correction for
this
error?

Thanks,

Hakan G

Here is some details about our system:

Client Side
OS: Windows Mobile 2003 4.21.1088
DB: SQL CE 2.0
Microsoft SQL Server CE (ssce20.dll) 2.00.4415.0
Microsoft SQL Server CE Client Agent (ssceca20.dll) 2.00.4415.0
Development Tools: VB.NET 2003
Service Pack: .NET Compact Framework 1.0 SP3

Server Side
OS: Microsoft 2003 SP1
Internet Information Services (INETINFO.EXE) 6.0.3790.1830
(srv03_sp1_rtm.050324-1447)
IIS Worker Process (w3wp.exe) 6.0.3790.1830 (srv03_sp1_rtm.050324-1447)
HW:IBM XSERIES_346 Intel(R) Xeon(TM) CPU 3.60GHZ (2CPU) 5,00 GB RAM DB:
SQL CE 2.0
DB:SQL Server Standart Edition 8.00.2039(SP4)

SQL CE Server 2.0
Microsoft SQL Server CE Server Agent (sscesa20.dll) 2.00.7331.0
Microsoft SQL Server CE Replication Provider (sscerp20.dll) 2.00.7331.0

Merge Replication Properties
--
status : 1
retention : 21
sync_mode : 1
allow_push : 1
allow_pull : 1
allow_anonymous : 1
centralized_conflicts : 1
priority : 100.0
snapshot_ready : 1
publication_type : 1
enabled_for_internet : 0
dynamic_filters : 1
has_subscription : 0
snapshot_in_defaultfolder : 1
alt_snapshot_folder : NULL

Merge Agent Profile:
parameter_name value
--
-BcpBatchSize 100000
-ChangesPerHistory 100
-DestThreads 4
-DownloadGenerationsPerBatch 500
-DownloadReadChangesPerBatch 500
-DownloadWriteChangesPerBatch 500
-FastRowCount 1
-HistoryVerboseLevel 1
-KeepAliveMessageInterval 300
-LoginTimeout 15
-MaxDownloadChanges 0
-MaxUploadChanges 0
-MetadataRetentionCleanup 1
-NumDeadlockRetries 5
-PollingInterval 60
-QueryTimeout 300
-SrcThreads 3
-StartQueueTimeout 300
-UploadGenerationsPerBatch 100
-UploadReadChangesPerBatch 100
-UploadWriteChangesPerBatch 100
-Validate 0
-ValidateInterval 60

You have a high-volumne replication scenario going on there and while I have not seen this specific error, I know you can increase the default thread pool used on the IIS server to broker all of this replication. There is a way to get -SrcThreads 3 to a larger number and I just saw it in the Books OnLine recently - I'll have a look and see if I can give you more details on this.

In the meantime perhaps Laxmi knows how to tune this setting off the top of his head.

-Darren

|||

For configuring SQL Mobile Server Agent on IIS: Here are typical knobs,

1) MAX_THREADS_PER_POOL - No.of threads to handle the incoming sync requests

2) CLEANUP_INTERVAL - Sync sessions cleanup

3) USAGE - Restricting the support only to RDA/Merge/Both

Darren, Sorry I did not get what is meant by SrcThreads.

Thanks,

Laxmi Narsimha Rao ORUGANTI, MSFT, SQL Mobile, Microsoft Corporation

merge replication error

Hi All,
I,m trying to pull a suscription for the first time so I create the database
when I create the suscription but I'm getting this error.
could not be propagated to the subscriber.
(Source: Merge Replication Provider (Agent); Error number: -2147201001)
Unable to replicate a view or function because the referenced objects or
columns are not present on the Subscriber.
(Source: SQLServer (Agent); Error number: 0)
Invalid object name 'dbo.Clients'.
(Source: SQLServer (Data source); Error number: 208)
Invalid object name 'dbo.EmailJobReports'.
(Source: SQLServer (Data source); Error number: 208)
Invalid object name 'dbo.EmailLinks'.
(Source: SQLServer (Data source); Error number: 208)
Invalid object name 'dbo.Employees'.
(Source: SQLServer (Data source); Error number: 208)
Why is not creating this tables?
Tks in advance for you help..
Johnny
Tks Paul,
I will put my views in a separate publication.
Johnny
"Paul Ibison" <Paul.Ibison@.Pygmalion.Com> wrote in message
news:044401c4d7c8$5d17d310$a601280a@.phx.gbl...
> JFB,
> on SQL Server 2005 you can specify the article processing
> order, but not on SQL Server 2000. To avoid this type of
> issue, you could put the views into a separate
> publication or use sp_addscriptexec to send over
> separately (not forgetting the permissions though).
> Rgds,
> Paul Ibison
>
|||Are these object present on the subscriber? Are these objects part of your
publication?
If they are not present on your subcsriber and are not part of your
publication you will have to manually script them out and deploy them to
your subscriber. Possibly by using sp_addscriptexec if all of your
subscriptions were deployed using a UNC.
you could also use a pre snapshot script to deploy these objects.
If they are not present on your subscriber and are part of your publication
they are not being created in the correct order.
Replication does order the article creation scripts correctly, however it
depends on sysdepends on the publisher which can be inacurate. Your best
option again is a pre snapshot command containing a script which will
precreate these objects for you.
Hilary Cotter
Looking for a SQL Server replication book?
Now available for purchase at:
http://www.nwsu.com/0974973602.html
"JFB" <jfb@.newSQL.com> wrote in message
news:OyqGUZ81EHA.3820@.TK2MSFTNGP11.phx.gbl...
> Hi All,
> I,m trying to pull a suscription for the first time so I create the
> database when I create the suscription but I'm getting this error.
> could not be propagated to the subscriber.
> (Source: Merge Replication Provider (Agent); Error number: -2147201001)
> ----
> Unable to replicate a view or function because the referenced objects or
> columns are not present on the Subscriber.
> (Source: SQLServer (Agent); Error number: 0)
> ----
> Invalid object name 'dbo.Clients'.
> (Source: SQLServer (Data source); Error number: 208)
> ----
> Invalid object name 'dbo.EmailJobReports'.
> (Source: SQLServer (Data source); Error number: 208)
> ----
> Invalid object name 'dbo.EmailLinks'.
> (Source: SQLServer (Data source); Error number: 208)
> ----
> Invalid object name 'dbo.Employees'.
> (Source: SQLServer (Data source); Error number: 208)
> ----
> Why is not creating this tables?
> Tks in advance for you help..
> Johnny
>
|||Tks for reply Hilary,
These objects are not in the suscriber and are part of the publication.
How can i create it in the correct order?
can you show me how to use a pre snapshot command?
Sorry but I'm new in replication.
Also Paul tell me that I can include the views in a different publication,
because I soon I get out the views the replication works but of course my
database is not complete.
Tks again
Johnny
"Hilary Cotter" <hilary.cotter@.gmail.com> wrote in message
news:OhWVHs81EHA.3504@.TK2MSFTNGP12.phx.gbl...
> Are these object present on the subscriber? Are these objects part of your
> publication?
> If they are not present on your subcsriber and are not part of your
> publication you will have to manually script them out and deploy them to
> your subscriber. Possibly by using sp_addscriptexec if all of your
> subscriptions were deployed using a UNC.
> you could also use a pre snapshot script to deploy these objects.
> If they are not present on your subscriber and are part of your
> publication they are not being created in the correct order.
> Replication does order the article creation scripts correctly, however it
> depends on sysdepends on the publisher which can be inacurate. Your best
> option again is a pre snapshot command containing a script which will
> precreate these objects for you.
> --
> Hilary Cotter
> Looking for a SQL Server replication book?
> Now available for purchase at:
> http://www.nwsu.com/0974973602.html
> "JFB" <jfb@.newSQL.com> wrote in message
> news:OyqGUZ81EHA.3820@.TK2MSFTNGP11.phx.gbl...
>
|||Note that in my case I don't have inter-view
dependencies, so this is possible. If you have views that
query other views, then the same issue may arise. In that
case you'll need to consider sp_addscriptexec.
Rgds,
Paul Ibison
|||Tks for u reply Paul,
How can I use sp_addscriptexec in my snapshot or publication?
You know I'm new and I dont know how to use this... I create the script for
the views in a sql file..
Regards
Johnny
"Paul Ibison" <Paul.Ibison@.Pygmalion.Com> wrote in message
news:082701c4d865$fbb42440$a601280a@.phx.gbl...
> Note that in my case I don't have inter-view
> dependencies, so this is possible. If you have views that
> query other views, then the same issue may arise. In that
> case you'll need to consider sp_addscriptexec.
> Rgds,
> Paul Ibison
>
|||The details are in BOL
(http://msdn.microsoft.com/library/default.asp?
url=/library/en-us/tsqlref/ts_sp_repl_6qzn.asp). It isn't
really added to the publication, but is a way of using
the replication 'plumbing' to run a script (using osql).
Rgds,
Paul Ibison SQL Server MVP, www.replicationanswers.com
(recommended sql server 2000 replication book:
http://www.nwsu.com/0974973602p.html)

Merge Replication Conflict Resolution Latency?

I've been having a hard time finding a definitive answer about merge
replication and any sort of latency on a transaction between a
subscriber and publisher. I apologize in advance for the lengthy post,
but it's difficult to really "cut to the meat" of this one
Here's my scenario: I'm developing a .NET CF application on a PocketPC
that stores data in a SQL CE database that merges with a SQL 2000
publisher when the PDA is docked. Generally speaking, what goes up is
properly reflected when it comes back down via a "is complete" row
filter that simply looks for a flag that is set to 1. For example:
-Publisher has 128 rows
-PDA has 128 rows, 28 rows have been "completed"
-Instantiate merge with publisher
-Result is that there are 128 rows on publisher, 28 are complete, row
filter disregards 28 complete records and sends down 100 records that
are still incomplete to PDA
This seems to work great. Now enter conflicts. I have the resolver set
such that the subscriber always wins. We've had several cases where the
server has been under moderate to heavy load, and there have been a lot
of conflicts. This was an actual occurrence:
-Publisher has 79 rows
-PDA has 79 rows, 69 rows have been "completed"
-Publisher had 79 rows *removed*
-Instantiate merge with publisher
-Result is that there are 69 conflicts at publisher, the subscriber wins
all, but there are still *79* rows on the PDA after merge finishes
-I wait a minute and try again, this time there are about 50 rows on the
subscriber
-I repeat this process several times over the course of 5 minutes until
finally, I am left with the correct number of records on the PDA (0)
My question is, is this normal behavior for the resolver? Are conflicts
generally allowed to be "handed off" and the merge allowed to complete
without reflecting their outcome? If so, is there any way to tune or
change this behavior?
This also has me concerned about a normal merge without any conflicts.
Would there ever be a case in which a merge would get out of sync in
this way? My understanding is that instantiating a merge will always
result in both the supplying subscriber and publisher ending up with the
same data at that moment. My merge agents are all using the default
profile.
Thanks a BUNCH for any help you can offer. I've had an impossible time
answering these questions and I'm a programmer before a DBA
-Mike
Mike wrote:

> -Publisher has 79 rows
> -PDA has 79 rows, 69 rows have been "completed"
> -Publisher had 79 rows *removed*
> -Instantiate merge with publisher
> -Result is that there are 69 conflicts at publisher, the subscriber wins
> all, but there are still *79* rows on the PDA after merge finishes
> -I wait a minute and try again, this time there are about 50 rows on the
> subscriber
> -I repeat this process several times over the course of 5 minutes until
> finally, I am left with the correct number of records on the PDA (0)
I apologize, I had my scenarios crossed. This should have read:
-Publisher has 79 rows
-PDA has 79 rows, 69 rows have been "completed"
-Publisher had all 79 rows *modified*
-Instantiate merge with publisher
-Result is that there are 69 conflicts at publisher, the subscriber wins
all and they are noted as resolved in the agent history, but there are
still *79* rows on the PDA after merge finishes
-I wait a minute and try again, this time there are about 50 rows on the
subscriber
-I repeat this process several times over the course of 5 minutes, each
time the number of rows at the subscriber decreases, until finally, I am
left with the correct number of records on the PDA (10)
Thanks
-Mike