Replication question

  • Hello all

    Is it possible to have two publications against one publisher and 2 subscriptions at the other end also going to the same database on the subscriber?

    We are trying to get round a problem with replicating large amounts of data and I am thinking if we can replicate in parallel this may halve the initialisation.

    Thanks

  • Where is your bottle neck?

    If it is the logreader the only way around this bottleneck is to create another database and publication; data could be fed into here using triggers, views or four part naming. This is because you can only have one logreader per database. It doesnt matter how many publications you have.

    If its just the distribution agent then yes spliting logically into multiple publications will work for you provided the articles are unique. You will have to bear in mind your transactions and make sure articles are group logically so the data gets applied in the correct order.

  • Also, have you considered initialising from a back up or using a no-sync initialisation?

  • Hi Mr Mystery 🙂

    The bottleneck is the fact that after an Axapta upgrade, which performs Drop and Creates rather than Alters, we find ourselves in a situation where we have to re-initialise our subscriptions. The snaphot for some of the larger publications are taking 45mins while the actual bulk copy is taking 2.5hours.

    We take backups using Litespeed so am unsure if it would be possible to re-initialise from backup?

  • It doesnt matter which method of backup used since this is manual backup/restore when initialising from backup.

    http://msdn.microsoft.com/en-us/library/ms147834.aspx

    Is the upgrade process out of your hands completely?

  • The upgrade in Microsoft Axapta is all done through the application using a new language some kind of cross between SQL and C# and is auto generated by the application and we have no control over it at all unfortunately.

  • Kwisatz78 (2/24/2012)


    The upgrade in Microsoft Axapta is all done through the application using a new language some kind of cross between SQL and C# and is auto generated by the application and we have no control over it at all unfortunately.

    I would look into intialising this from a backup then as its something that can potentially be a lot quicker with a bit of preparation; such as setting up a logship before the update occurrs etc and using this as my backup initialisation.

  • Ok cool thanks for that, I will check out using log shipping with replication.

    Thanks

Viewing 8 posts - 1 through 8 (of 8 total)

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