Clustering WITH log shipping

  • I had a problem in the past with clustering where the RAID controller card died and corrupted several databases, requiring us to go from our last backup. Now my plan is to use clustering for high availability along with Log shipping to a separate box with its own disks for DR. If my controller or the RAID array unit itself goes bad, I want to be able to access my backups. Therefore, I'd like to also ship logs to another server.

    SQL Server clustered does not seem to support mounted drives (am I wrong?). Therefore, how can I tell SQL Server to ship my logs to another machine? Do I have to put my log ship server into the cluster now?

    thanks

    -gary

  • Hi,

    you should never use mapped drives for logshipping, because the drive mapping lives in the user profile.

    use UNC or ip address for the shipping:

    \\servername\share

    or

    \\172.16.1.1\share

    A basic introduction how to manually set up logshipping is

    http://www.sql-server-performance.com/sql_server_log_shipping.asp

    or you can use maintanance plans in EM (if you like black boxes...)

    regards karl

    Best regards
    karl

  • This article is interesting. I used the Maintenance Wizard w/ EM so my situation is slightly different. Anyway, the problem stemmed from the fact that when setting up log shipping using the Maintenance Plan wizard, it required me to select a known fixed drive (or mounted drive). This causes some difficulty when using a cluster because the cluster can only access devices within the cluster. The resolution to this was actually fairly simple. My view of this was that I needed to create a file share on the Log Ship server and ship the logs to that server. I had to alter this vision slightly and I created a File Share resource on the Cluster, so that the log ship server now has to grab the files from the main cluster array. Not originally how i'd planned it, but not a big deal. Thanks for your response.

  • You might not be able to prevent database corruption due to controllers malfunction using logshipping because corruption will propagate to the DR server via the transaction logs restore.

    The logshipping will be usefull if databases get corrupted due to other reasons or system databases get corrupted or both SQL Server installations in the cluster stop working.

Viewing 4 posts - 1 through 3 (of 3 total)

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