Help! DataSet.Merge causing duplicates

[Derrick]

Hello;

I've got a problem where DataSet.Merge is causing duplicate records,
and then throwing a constraint failure exception because of duplicate
primary keys.

I've seen a similary problem discussed here where the key fields were
set to Auto-increment integers. I am not using this - I'm using GUIDs
as primary keys.

My scenario: I'm doing a (supposedly) simple remoting app. Note -
there is no database server, I'm only using the dataset as a
collection, not to do any SQL. Instances of the same typed dataset
exist on both sides of the remoting channel (I'll call one "local",
the other "remote"). On the remote side, when a row in any table is
changed, I call the DataSet.GetChanges() method to extract those
changes. I then send this "sub-dataset" from the remote side to the
local side, where it is to be merged with it's copy, thus updating
it. This is where I get the exception.

I can look at the data before the merge, and everything seems to line
up. If I only modify one row, I can look at both copies and see that
row's state listed as "Unchanged" (with the old data) in the local
copy, and "Modified" (with the new data) in what was received from the
remote side. I can compare the bytes of the GUID primary key and see
they are equal. I can see the field whose data has changed. From the
description given on MSDN, I seem to have all the criteria for a
successful merge (http://msdn2.microsoft.com/en-us/library/
4swwh51k.aspx). However, when I look at the local data after the
merge blows up on the constraint exception, I can see the duplicate
field in my table, with both the old & new data.

This has me stumped. I could write my own merge function, but why?
If anyone can shed light on this, I'd appreciate it!

Derrick

 

[openshac]

Have you tried calling the dataset method AcceptChanges() before calling Merge() ?
This worked for me when I was getting duplicate rows.

 

[Richard Cooke]

Did you discover how to fix this? I have exactly the same problem with having a duplicate row created after merging and updated dataset.



Derrick wrote:

Help! DataSet.Merge causing duplicates
16-Feb-08

Hello

I've got a problem where DataSet.Merge is causing duplicate records
and then throwing a constraint failure exception because of duplicat
primary keys

I've seen a similary problem discussed here where the key fields wer
set to Auto-increment integers. I am not using this - I'm using GUID
as primary keys

My scenario: I'm doing a (supposedly) simple remoting app. Note
there is no database server, I'm only using the dataset as
collection, not to do any SQL. Instances of the same typed datase
exist on both sides of the remoting channel (I'll call one "local"
the other "remote"). On the remote side, when a row in any table i
changed, I call the DataSet.GetChanges() method to extract thos
changes. I then send this "sub-dataset" from the remote side to th
local side, where it is to be merged with it's copy, thus updatin
it. This is where I get the exception

I can look at the data before the merge, and everything seems to lin
up. If I only modify one row, I can look at both copies and see tha
row's state listed as "Unchanged" (with the old data) in the loca
copy, and "Modified" (with the new data) in what was received from th
remote side. I can compare the bytes of the GUID primary key and se
they are equal. I can see the field whose data has changed. From th
description given on MSDN, I seem to have all the criteria for
successful merge (http://msdn2.microsoft.com/en-us/library
4swwh51k.aspx). However, when I look at the local data after th
merge blows up on the constraint exception, I can see the duplicat
field in my table, with both the old & new data

This has me stumped. I could write my own merge function, but why
If anyone can shed light on this, I'd appreciate it

Derrick

Previous Posts In This Thread:

Help! DataSet.Merge causing duplicates
Hello

I've got a problem where DataSet.Merge is causing duplicate records
and then throwing a constraint failure exception because of duplicat
primary keys

I've seen a similary problem discussed here where the key fields wer
set to Auto-increment integers. I am not using this - I'm using GUID
as primary keys

My scenario: I'm doing a (supposedly) simple remoting app. Note
there is no database server, I'm only using the dataset as
collection, not to do any SQL. Instances of the same typed datase
exist on both sides of the remoting channel (I'll call one "local"
the other "remote"). On the remote side, when a row in any table i
changed, I call the DataSet.GetChanges() method to extract thos
changes. I then send this "sub-dataset" from the remote side to th
local side, where it is to be merged with it's copy, thus updatin
it. This is where I get the exception

I can look at the data before the merge, and everything seems to lin
up. If I only modify one row, I can look at both copies and see tha
row's state listed as "Unchanged" (with the old data) in the loca
copy, and "Modified" (with the new data) in what was received from th
remote side. I can compare the bytes of the GUID primary key and se
they are equal. I can see the field whose data has changed. From th
description given on MSDN, I seem to have all the criteria for
successful merge (http://msdn2.microsoft.com/en-us/library
4swwh51k.aspx). However, when I look at the local data after th
merge blows up on the constraint exception, I can see the duplicat
field in my table, with both the old & new data

This has me stumped. I could write my own merge function, but why
If anyone can shed light on this, I'd appreciate it

Derrick


Submitted via EggHeadCafe - Software Developer Portal of Choice
WPF Report Engine, Part 4
http://www.eggheadcafe.com/tutorial...45-8b37cb7f3186/wpf-report-engine-part-4.aspx

 

[Hans Andersen]

Make sure the key is a primary key and not just a key. It took me hours before I noticed that I'd forgotten to tick the "Primary key" checkbox in the data-table designer. Once I checked it my merge worked perfectly.