José Moreira Neto | Microsoft SQL Server Escalation Services Problem 1: batch updates 1) A single update command can generate a transaction with thousands of commands. For example if you execute an UPDATE statement that updates 10K rows on the publisher, up to 20K commands can be send over to the subscriber. On the publisher
Read More »
Sateesh Yele | Microsoft SQL Server Support I would like to share information about an interesting Replication issue I recently worked. Customer was getting unusual error messages during synchronization. Symptom: Merge agent is failing with the following error message when validating large tables with checksum. The merge process was unable to perform data validation on
Read More »
Sateesh Yele | Microsoft SQL Server Support Peer-2-Peer Replication Wizard provides no options to enable table partition switching. There are 2 ways to enable this feature. 1) create the publication on ALL the peers using TSQL sp_addpublication 2) create the subscriptions using the user interface — or — 1) create the publication using the Replication
Read More »
Contributed by: Karol Korasadowicz Depending on type of data being replicated and network throughput adjusting the PacketSize parameter for the Distribution Agent may increase your overall Transactional Replication performance. -PacketSize packet_size Is the packet size, in bytes. The default is 4096 (bytes). http://msdn.microsoft.com/en-us/library/ms147328.aspx To see which setting has most impact, create test scenario which inserts/updates
Read More »
As we know, Transactional Replication moves data from the Publisher to the Distributor then to the Subscribers. When working, data flows near real-time through both of these steps. When problems occur which interrupt this flow, latency is introduced. The challenge is to find which step is contributing to the overall latency. Tracer Tokens help you
Read More »