Does the Log shipping configuration breakes in a a/p cluster setup when a failover occurs?

  • please provide me the BOL link which explains How to configure log shipping in clustered sql server 2005

  • What error is being returned by the copy job?

    Ensure that the network name your are using to get to the share on the primary is the Windows virtual name for the cluster.

    Start here in BOL: ms-help://MS.SQLCC.v9/MS.SQLSVR.v9.en/udb9/html/30e06a7d-75e9-44e2-bca3-b3b0c4a33f61.htm

    or go to the index and search for "failover clustering"



    Scott Duncan

    MARCUS. Why dost thou laugh? It fits not with this hour.
    TITUS. Why, I have not another tear to shed;
    --Titus Andronicus, William Shakespeare


  • Thanks,

    The copy job, to the secondary failing with below error:

    Error: The network name cannot be found

    It would be great help if you can point me to the link How to configure Log shipping in Clustered SQL Server 2005, NOT the failover cluster setup.

    COuld you please point me to How to configure Log Shipping In Clustered SQL Server 2005 in BOL. I'm looking for this from past 3 days but I did not find any

    Please help me out

  • There is nothing special about configuring log shipping in a clustered environment compared to a standalone instance. They are setup/configured the same way.

    Here is your starting point: http://msdn.microsoft.com/en-us/library/ms188698(SQL.90).aspx

    Your problems appear to be related to clustering and are not specific to log shipping. If you need a drive to be available in the GUI so you can setup the transaction log backups, you need to add a disk resource to the SQL Cluster Group. Once you have the disk resource created, then you make SQL Server dependent upon that disk and it will be available to SQL Server in the GUI.

    If you do not make SQL Server dependent upon the disk resource, then the GUI tools will not 'see' the drive. It is still available and accessible by drive letter because it exists in the same cluster group.

    If you put the disk resource into a separate cluster group - it is still going to be available to SQL Server, but not with the drive letter. You have to reference it using the UNC path.

    This has all been said before...I am not sure what else can be said.

    Jeffrey Williams
    “We are all faced with a series of great opportunities brilliantly disguised as impossible situations.”

    ― Charles R. Swindoll

    How to post questions to get better answers faster
    Managing Transaction Logs

  • Thanks for all your help...

    and one more question: does we need to give the backup path as

    \\VirtualserverName(instead of Node name)\Z:\logshipping_backups_out in Log shipping wizard backup settings page?

  • klnsuddu (6/29/2009)


    Thanks for all your help...

    and one more question: does we need to give the backup path as

    \\VirtualserverName(instead of Node name)\Z:\logshipping_backups_out in Log shipping wizard backup settings page?

    The UNC will not include the drive letter, replaces the drive letter. You may want to do some research on UNC's. Windows Help, not SQL Server Help, is probably the best place to start.

  • \\VirtualserverName(instead of Node name)\logshipping_backups_out in Log shipping wizard backup settings page?

    here logshipping_backups_out is the file share created in the cluster adminstrator..

    does this makes sense?

  • klnsuddu (6/29/2009)


    Thanks for all your help...

    and one more question: does we need to give the backup path as

    \\VirtualserverName(instead of Node name)\Z:\logshipping_backups_out in Log shipping wizard backup settings page?

    If the drive is in the same cluster group as SQL Server, then you can just use Z:\logshipping_backups_out. If the drive is in a different cluster group, then you need to reference the drive using \\{cluster group ip name of other cluster group}\Z$\logshipping_backups_out.

    Each cluster group should have it's own network name - which is what you would need to use to reference a drive in that cluster group. Well, that is as long as you have permissions to that drive.

    Jeffrey Williams
    “We are all faced with a series of great opportunities brilliantly disguised as impossible situations.”

    ― Charles R. Swindoll

    How to post questions to get better answers faster
    Managing Transaction Logs

  • klnsuddu (6/29/2009)


    Thanks for all your help...

    and one more question: does we need to give the backup path as

    \\VirtualserverName(instead of Node name)\Z:\logshipping_backups_out in Log shipping wizard backup settings page?

    If the drive is in the same cluster group as SQL Server, then you can just use Z:\logshipping_backups_out. If the drive is in a different cluster group, then you need to reference the drive using \\{cluster group ip name of other cluster group}\Z$\logshipping_backups_out.

    Each cluster group should have it's own network name - which is what you would need to use to reference a drive in that cluster group. Well, that is as long as you have permissions to that drive.

    Jeffrey Williams
    “We are all faced with a series of great opportunities brilliantly disguised as impossible situations.”

    ― Charles R. Swindoll

    How to post questions to get better answers faster
    Managing Transaction Logs

  • \\VirtualserverName(instead of Node name)\logshipping_backups_out in Log shipping wizard backup settings page?

    here logshipping_backups_out is the file share created in the cluster adminstrator..

    finallay, the bove settings worked, even after moving to passive node also, the backup, copy and restore jobs working....

    thank you....

  • \\VirtualserverName(instead of Node name)\logshipping_backups_out in Log shipping wizard backup settings page?

    here logshipping_backups_out is the file share created in the cluster adminstrator..

    finallay, the bove settings worked, even after moving to passive node also, the backup, copy and restore jobs working....

    Is that necessary and correct method of giving Virtual server name in the backup path?

    Its working for me but I need some justification though.

  • You need to reference the virtual name so that you are not tied to a single node in your cluster.

    For example, say you have clusternode1 & clusternode2 in an active/passive cluster called cluster1. You have a clustered file share fileshare1. The current active node is clusternode1.

    You setup log shipping (or anything else that refers to the file share) and use \\clusternode1\fileshare1. Things work OK.

    Then the cluster fails over to clusternode2. Then, any references to \\clusternode1\fileshare1 will no longer work, as fileshare1 is no longer associated with clusternode1.

    If you use \\cluster1\fileshare1, then it doesn't matter if the cluster fails over - the cluster1 name will redirect to whatever the active node is.

    Bear in mind this is for active/passive scenario. I don't know how it would work for active/active.

    The point is, when accessing cluster resources, use the cluster virtual name.



    Scott Duncan

    MARCUS. Why dost thou laugh? It fits not with this hour.
    TITUS. Why, I have not another tear to shed;
    --Titus Andronicus, William Shakespeare


  • Thank you,

    I'm in good shape now. Because we have Active\passive cluster and I'm using Virtual server name in the backup path and log shipping is working.

    Now I'm testing Failover and Role switch...

    Thanks

  • For example, say you have clusternode1 & clusternode2 in an active/passive cluster called cluster1. You have a clustered file share fileshare1. The current active node is clusternode1.

    You setup log shipping (or anything else that refers to the file share) and use \\clusternode1\fileshare1. Things work OK.

    Then the cluster fails over to clusternode2. Then, any references to \\clusternode1\fileshare1 will no longer work, as fileshare1 is no longer associated with clusternode1

    Interesting (weird) thing I'm facing here is:

    Here the Backup job is NOT failing even if we give \ode1(instead of virtual server name) after the sql group fails to node2. The COPY job(which copies the tlog to secondary server) is Failing.

    But when we use Virtual server name , then both jobs working fine..

    Any Idea..why the Backup job is NOT failing even if we give node name in the backup path and why copy job is failing??

  • Interesting.

    Not sure how clustering handles things under the hood, but from what I've seen, a cluster resource (file share at least) can be connected to by either the virtual name or the hostname of the node the resource is running from (the active node). Trying to do the same with the passive node is the same as trying to connect to a share that doesn't exist on any given server. Like if I try connecting to \\mycomputer\myfileshare - I don't have a share called "myfileshare" on my computer, so it will fail to connect. Same for trying to connect to a clustered file share using the passive node hostname - the share doesn't exist on that node, so connection attempts will fail.

    Not sure why the backup doesn't have this problem - maybe because it is using Z:\folder instead of a UNC?

    I can't think of a better way to explain this. I'm no clustering expert, just know what I have seen & encountered along the way.



    Scott Duncan

    MARCUS. Why dost thou laugh? It fits not with this hour.
    TITUS. Why, I have not another tear to shed;
    --Titus Andronicus, William Shakespeare


Viewing 15 posts - 16 through 30 (of 37 total)

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