Does The Actual Update Statement Of A Table Affect Merge Replication?
Dec 8, 2006
Suppose your using Merge Replication.
2 Users in 2 locations issue updates to the same table. 1 updating 1 column and the other updating another column. Now in reality the actual Stored Procedure issuing the update statement is passed in all the possible columns that could change and builds an update statement that updates all columns even the ones that havent changed.
Will this break Merge Replications conflict tracking? Or does SQL Server 2005 Merge Replication pickup that in reality the 2 updates only in reality changed the values in 2 columns.
Could a simple update statement on a user database ever caused space usage in tempdb? Â Assuming the update statement fires no triggers and not using any temp tables?
IE:
User DatabaseA Update TableX Set col1 = X
Reason I ask is tempdb filled up and the only thing I could see running at that time was the update statement.
In a website I'm working on I need to be able to put all of the user tables in a database in a dropdownlist.
Another dropdownlist then will autopopulate itself with the names of all the columns from the table selected in the first dropdownlist.
So, what I need to know is: is there a sql statement that can return this type of information?
Example:
Table Names in Database: Customers, Suppliers
Columns in Customers Table: Name, Phone, Email, Address
I click on the word "Customers" in the first dropdownlist.
I then see the words "Name", "Phone", "Email", "Address" in the second dropdownlist.
I'm sure you all know this (but I'll say it anyways): I could hardcode this stuff in my code behind file, but that would be really annoying and if the table structure changes I would have to revise my code on the webpage. So any ideas on how to do this the right way would be really cool.
i have a little question. Is it possible you can't perform an insert statement on a table wich is replicated with merge replication?
I set the replication up and everything works fine, but if i want to perfom an insert statement on the table, i get an error that the values i want to add aren't the same as the one in the table.
I know that merge replication creates a new column and I think that's the problem.
Can someone help me solve this or confirm that you can't perform an insert statement on a replicated table?
Hi, I'am using several SQL 2000 servers with different working merge-replications. Actually I wrote a .NET assembly which makes updates to key values in one of the replicated tables:
Ex: update ReplTable set key = -key where key between (1 and 100)
Loop ...... update ReplTable set key = NewKey where key = .... end of loop
Intention is to resort the datarows. So I made two updates to every keyvalue in a very short time. At the ConflictViewer I got messages like:
"The row was inserted at 'ServerX' but could not be inserted at 'ServerY'. Violation of PRIMARY KEY constraint 'PK_xxxx. Cannot insert duplicate key in object 'ReplTable'. "
The replication did only deliver one (the last) update - this makes the violation. Is there a possibility to make the mergeagent known that I wanted to deliver every transaction? Is there a parameter in the agents profile?
In the following t-sql 2012 merge statement, the insert statement works but the update statement does not work. I know that is true since I looked at the results of the update statement:
Merge TST.dbo.LockCombination AS LKC1 USING (select LKC.comboID,LKC.lockID,LKC.seq,A.lockCombo2,A.schoolnumber,LKR.lockerId from [LockerPopulation] A JOIN TST.dbo.School SCH ON A.schoolnumber = SCH.type
[Code] ...
Thus can you show me some t-sql 2012 that I can use to make update statement work in the merge function?
I am trying to create Replication Topology (Merge Replication) like below.
Subscriber1 --> Publisher <-- Subscriber2.
I have created both subscribers with Subscription Type as Server with Priority as 75. I am updating the Column A of Row_10 in Subscriber1 on time say 11 am. After i am running the Starting synchronizing agent from Subscriber1. The value propagated to Publisher now publisher contains the latest value in Column A. Uptonow the Subscriber2 is not synchronized with Publisher.
Now in Subscriber2 also Column A of Row_10 is updated say at 11.10 am. Actually now Publisher contains the value from Subscriber1 for that Column and in Subscriber2 we have the same column updated.
Now i am running the Synchronization in Subscriber2, i am getting the result which is not expected. Here Publisher's value is propagated to Subscriber2. But as per real scenario Subscriber2 has the Latest value which is updated on 11.10 AM.
I don't know what am i missing here. Actually merge replication should see the time stamp and it has to decide winner. But here it always considers publisher as a winner and puts the data to Subscriber.
I have got a business logic update conflict handler working, but I have had to work round what appears to be a bug.
Please can someone confirm if this is indeed a bug €“ and if it is a known bug?
My conflict handler needs to take some columns from the publisher row and some from the subscriber row in the event of conflict.
I can quite happily generate a custom dataset which contains the winning row that I want €“ I can see that because I can step through the conflict handler with debug when a conflict occurs.
However, just returning ActionOnUpdateConflict.AcceptCustomConflictData from the UpdateConflictsHandler method does not set the publisher and subscriber columns correctly. I end up with different values on the two databases.
I have found that the only way to get the correct rows on both publisher and subscriber is to create a new ADO connection to the publisher and actually perform an update €“ updating all the modified columns. This now works reliably in my testing.
Fortunately, due to business rules the frequency of update conflicts are likely to be very infrequent, but I would very much like to avoid having to do the €˜unnecessary€™ update.
Notes:
I am using column level tracking €“ but I have seen the problem with row level tracking too I have mainly been using SP1 but have repeated the test on a configuration using the SP2 CTP €“ and the problem occurs there too The problem is not due to complex logic in my code. If the method just sets customDataSet = publisherDataSet.Copy and then returns ActionOnUpdateConflict.AcceptCustomConflictData, the changed and winning publisher values are not sent to the subscriber
I have recently setup a transactional replication in MS SQL 2000. After setting up the replication the clients TempDB grew by almost 60GB. Now the client is Blaming me for the TempDB GROWTH and saying that its because of the replication being setup i tried to convince them but they are not satisfied yet. Can anybody please tell me does replication cause the tempdb to grow. If yes then how. can u suggest any good link for getting to know the internal working of SQL Server replication????
I have created a Dynamic Merge statement SCD2 Store procedure , which insert the records if no matches and if bbxkey matches from source table to destination table thne it updates old record as lateteverion 0 and insert new record with latest version 1.
I am getting below error when I ahve more than 1 bbxkey in my source table. How can I ignore this.
BBXkey is nothing but I am deriving by combining 2 columns.
Msg 8672, Level 16, State 1, Line 6
The MERGE statement attempted to UPDATE or DELETE the same row more than once. This happens when a target row matches more than one source row. A MERGE statement cannot UPDATE/DELETE the same row of the target table multiple times. Refine the ON clause to ensure a target row matches at most one source row, or use the GROUP BY clause to group the source rows.
We have a big problem here with merge replication, specifically whenever a schema change occurs. We are replicating schema changes, and triggers/stored procedures. The example is that we changed about 150 stored procedures and about 30 triggers. This is then replicated to the subscriber database (which is also a merge publisher for further remote systems that were offline at the time) over a 10Mb link - hardly low bandwidth. However, the replication process takes about an hour and a half - considering the SQL on the primary server took less than a minute to run this is a big suprise.
We've run a trace to see if we can identitfy what is going on. There seems to be a great number of calls to sp_MSunmarkschemaobject - we are still waiting for a trace to complete to fully analyse this however it looks like it calls this repeatedly for every stored procedure in the database. We are currently re-testing to one of the remote servers with the merge agent set to the slow profile (not much hope that this will alter the poor performance).
This task looks to be excessive - and certainly does not seem to function in a sensible manner. Has anyone else had similar issues or have any suggestions. This is very infuriating as it means that the servers are effectively offline for a minimum of and hour and a half (in fact the remoter servers take over 4 hours !).
Problem Summary: Merge Statement takes several times longer to execute than equivalent Update, Insert and Delete as separate statements. Why?
I have a relatively large table (about 35,000,000 records, approximately 13 GB uncompressed and 4 GB with page compression - including indexes). A MERGE statement pretty consistently takes two or three minutes to perform an update, insert and delete. At one extreme, updating 82 (yes 82) records took 1 minute, 45 seconds. At the other extreme, updating 100,000 records took about five minutes.When I changed the MERGE to the equivalent separate UPDATE, INSERT & DELETE statements (embedded in an explicit transaction) the entire update took only 17 seconds. The query plans for the separate UPDATE, INSERT & DELETE statements look very similar to the query plan for the combined MERGE. However, all the row count estimates for the MERGE statement are way off.
Obviously, I am going to use the separate UPDATE, INSERT & DELETE statements. The actual query plans for the four statements ( combined MERGE and the separate UPDATE, INSERT & DELETE ) are attached. SQL Code to create the source and target tables and the actual queries themselves are below. I've also included the statistics created by my test run. Nothing else was running on the server when I ran the test.
Server Configuration:
SQL Server 2008 R2 SP1, Enterprise Edition 3 x Quad-Core Xeon Processor Max Degree of Parallelism = 8 148 GB RAM
SQL Code:
Target Table: USE TPS; IF OBJECT_ID('dbo.ParticipantResponse') IS NOT NULL DROP TABLE dbo.ParticipantResponse;
Hi,How to add new table to merge replication ?When I do this in Enterpriise Manager i got an error:Server: Msg 2758, Level 16, State 1, Procedure sp_addmergearticle,Line 466RAISERROR could not locate entry for error 20085 in sysmessages.What is wrong ?Thank for any helpLuk
I have a table used in Merge Replication. If i add a new column in that table it gives me an error for tables in which either this table is a key or this table has key from other tables "Error validating the default for column 'rowguid'." Also for views i'm getting an error "View 'dbo.vw_TestCaseBasic': indexes and schema binding will be removed."
If i proceed further i get an error and m not able to add the column. "It is invalid to drop the default constraint on the rowguid column that is used by merge replication. The schema change failed during execution of an internal replication procedure. For corrective action, see the other error messages that accompany this error message. The transaction ended in the trigger. The batch has been aborted."
I tried with success to replicate a table_test_repl from my local server to an other server.
I used only one table for testing.
After that, i can't drop the 2 tables a_test_repl (on my local server and the other : I've got a message 'cannot drop the table because it is published for replication'
What can I do? I need absolutely to drop these tables to keep a clean model and I can't drop the database and try again.
I manage complete online system for two locations, connected via dialup lease line. I have to used merge replication for the same data should be at both locations. It is working fine, I don’t like disturbe it but current requirement is addition of two column in one table which one as also use in merge replication.
How I can add aforesaid columns in table x, column name c1 numeric , c2 character?
One of the table that is in a merge replication somehow is missing anindex. Strangely, only the table in one of the subscriber of the mergereplication is missing the index; another subscriber and the publisherof the merge replication don't have this problem.How should I add the missing index back to that table? My understandingis that making structural change on a table that is inmerge-replication is different from making change on a table that isnot merge-replicated. For example, when we need to add a column into atable that is being merge-replicated, we must add the column bychanging the attributes in the properties of the published article(table) instead of simply using CREATE INDEX command. I am wonderingwhether there is a similar restriction on adding an index onto a tablethat is merge replicated.I have already added the index back to the table anyway. I am askinghere just in case doing this may get me into a problem later on.Thanks in advance for any info.Jay Chan
Can someone point me to a resource for Table Design Considerations for Merge Replication? I have an ASP.Net/SQL2K5 app that I need to run on disconnected machines, then allow dfor data sync through merge replication. I assume that the first step is getting my tables indexed in a replication friendlt manner?
Many Thanks to anyone who can point me in the right direction!
I faced with strange problem. From time to time, during syncronization I get the following errors at the subscriber:
ERROR: -2147200992
SOURCE: Merge Replication Provider
TEXT: The merge process could not store conflict information for article 'Products'. Check the publication properties to determine where conflicts records are stored. When troubleshooting, restart the synchronization with verbose history logging and specify an output file to which to write.
ERROR: 2812
SOURCE: OFF1SQLEXPRESS TEXT: Could not find stored procedure 'MSmerge_cft_sp_49D28CEAFF8A4ACC5C5462B5027D4E4D'.
ERROR: -2147199481
SOURCE: Merge Replication Provider
TEXT: The Merge Agent was unable to update information about the last synchronization at the Subscriber. Ensure that the subscription exists at the Subscriber, and restart the Merge Agent.
After I refresh the snaphot at the publisher, these errors disappear and the syncronization is completed without any problem. And I can sync easy again week or two..
What is the problem? What is the right solution?
Yep, I can continue to refresh the snapshot each time when the error appears, but.. maybe there is another approach?
Hi,I have table with three columns as belowtable name:expNo(int) name(char) refno(int)I have data as belowNo name refno1 a2 b3 cI need to update the refno with no values I write a query as belowupdate exp set refno=(select no from exp)when i run the query i got error asSubquery returned more than 1 value. This is not permitted when thesubquery follows =, !=, <, <= , >, >= or when the subquery is used asan expression.I need to update one colum with other column value.What is the correct query for this ?Thanks,Mani
During merge replication initialization from SQL Server 2005 to SQL Server CE 3.5 on Windows Mobile 5.0
I am getting the following errors:
On Device {"The table name is not valid. [ Token line number (if known) = 1,Token line offset (if known) = 21,Table name = COLLECTION ]"} base {System.SystemException}: {"The table name is not valid. [ Token line number (if known) = 1,Token line offset (if known) = 21,Table name = COLLECTION ]"} HResult: -2147217900 Message: "The table name is not valid. [ Token line number (if known) = 1,Token line offset (if known) = 21,Table name = COLLECTION ]" NativeError: 26100 Source: "Microsoft SQL Server Compact"
{System.Data.SqlServerCe.SqlCeError: The SQL statement failed to execute. If this occurred while using merge replication, this is an internal error. If this occurred while using RDA, then the SQL statement is invalid either on the PULL statement or on the SubmitSQL statement. [ SQL statement = ALTER TABLE [dbo].[COLLECTION] ADD CONSTRAINT [PK_COLLECTION] PRIMARY KEY CLUSTERED ( [SERIAL_NUMBER], [STORE_CODE], [COLLECT_DATE] )WITH PAD_INDEX , FILLFACTOR = 90 ]} Source: "Microsoft SQL Server Compact"
Server Logged: ... 2008/04/20 20:03:16 Thread=EA0 RSCB=5 Command=SCHK Hr=00000000 SyncCheck responding 0 2008/04/20 20:03:17 Thread=B90 RSCB=5 Command=OPNR Hr=00000000 End Of Data Set 1 2008/04/20 20:03:17 Thread=B90 RSCB=5 Command=OPNR Hr=00000000 Responding to Fetch, bytes = 50259 2008/04/20 20:03:17 Thread=970 RSCB=5 Command=CLOS Hr=00000000 Total Compressed bytes out = 50259 2008/04/20 20:03:17 Thread=970 RSCB=5 Command=CLOS Hr=00000000 Total Uncompressed bytes out = 390254 2008/04/20 20:03:17 Thread=970 RSCB=5 Command=CLOS Hr=00000000 Removing this RSCB 0 2008/04/20 20:03:19 Hr=00000000 Compression Level set to 1 2008/04/20 20:03:19 Hr=80004005 ACK:Error for DistributorSessionID = 80 2008/04/20 20:03:19 Hr=80004005 The SQL statement failed to execute. If this occurred while using merge replication, this is an internal error. If this occurred while using RDA, then the SQL statement is invalid either on the PULL statement or on the SubmitSQL statement. [,,,SQL stateme 28560
The same tables in a different on a different SQL box (test) replicate fine, but on this box (production) they do not. Any idea what would be the difference in the two environments?
I suspect that the "dbo." part in the "alter table" is what it is messing up on... how can I tell SQL CE not to include that schema prefix? Thanks
I am new to use MERGE statement. The MERGE cannot find any match Cardnumber in the target table.  It inserts row into an existing row on the target table causing SQL rejected with duplicate key not allowed. The CardNumber is defined as a primary key on the target table with no duplicate allowed. Below snippet stop when MERGE insert a row exists on the target. The source table contains multiple rows with the same Cardnumber because it is a transactional table with multiple redemptions.Â
If MERGE cannot handle many (source) to one (target) relationship, what other method that I can change to in order to update the target GiftCard table which keeps track of gift card balance?Â
Below is the error message:
Msg 2627, Level 14, State 1, Line 5 Violation of PRIMARY KEY constraint 'PK_GiftCard'. Cannot insert duplicate key in object 'dbo.GiftCard'. The duplicate key value is (63027768).
declare tableName table ( uniqueid int identity(1,1), id int, starttime datetime2(0), endtime datetime2(0), parameter int )
A stored procedure has new set of values for a given id. Sometimes the startime and endtime are the same, in which case I update the value of parameter. Sometimes I add a new time range (insert statement), and sometimes I delete a time range (delete statement).
I had a question on merge, with insert, delete and update and I got that resolved. However I have a different question regarding performance of the merge statement.
If my target table has hundreds of millions of records and I want to delete/update/insert a handful of records, will SQL server scan the entire target table? I can't have:
merge ( select * from tableName where id = 10 ) as target using ...
and I can't have:
merge tableName as target using [my query] as source on source.id = target.id and source.starttime = target.startime and source.endtime = target.endtime where target.id = 10 ...
This means I cannot filter the set of rows in the target table to a handful of records where id = 10.
With merge/insert statements ...Is DISTINCT best way to handle problem of source table containing duplicate rows, along with WHERE NOT IN statement? the source dataset is large and having to do DISTINCT and further filtering is taxing on the ETL.
I have a table which is updated daily using a MERGE statement. As records are insert, updated and deleted, I am saving the OUTPUT from the MERGE statement into a history table with a timestamp and action$ column appended to the record.
Using this history table, I'd like to rebuild the data based on specific past date. I was able to create a stored procedure that inspects each record in the history table and apply it to the data in a temp table. The stored procedure solution uses multiple queries to rebuild the data at a point in time. I was curious if there was an easier and more efficient solution using a table function.
I have a view in SQLServer 2005. It took 30 sec. to finish. Then I deleted 4500 records from one table that is used in view. It took 90 sec. to finish now. I did a comparison on Actual Execution Plan between before I deleted data and after I deleted data, they are almost same, only different is Actual Number Rows become less after deleted data. So, I wonder why data become less but time become more. When I look closely on the Actual Execution Plan, the ridiculous thing is, there are only Estimated Operation Cost on each step, no Actual Operation Cost. I guess there are something wrong with optimizer because reuse same Execution Plan, but how can I tell which step wrong without Actual Operation Cost.
1) As per my project requirement, if the changes in SSAS cube are approved, they should be committed back to the actual SQL Server 2012 tables. Is that possible, if yes how?
2) For rolling back to original data I truncate the relevant writeback table and process the cube.
