Transactional Replication - Inner workings

  • Hi all

    Just wondering how the inner working of replication happens. eg. You have server A replication to Server B. (transactional replication). You have a pull replication setup on Server B with the Distribution db on Server A along with the publication

    Scenario:

    You have a job that imports data into server A at 5am (runs for 2 hours), another job at 6am (runs for 2 hours) and one more at 7am (runs for 2 hours). The data is being imported into SQL by these jobs and they getting transferred to Server B via replication. (Also please don't mention replication monitor, cause in my experience its not very accurate!!!). How does the data get committed to server A with multiple jobs running and how does the data transfer itself to server B. Is this done by means of a que on the distribution? how does the log reader identify which records to read first and which to transfer first on the publication? Are the transactions grouped into files? if it is, where are these files held and what order are they transferred?

    I know alot of questions, but any positive feedback will be highly appreciated!

    Thanks!!!

  • How about having a look at the info published by Microsoft on how replication works. There is stacks of info available in Book Online, Technet and MSDN. Perhaps start with "How Transactional Replication Works" (http://msdn.microsoft.com/en-us/library/ms151706(v=sql.100).aspx)

Viewing 2 posts - 1 through 1 (of 1 total)

You must be logged in to reply to this topic. Login to reply