Always On Issue 2016

  • Always On scenario:

    Listener

    Node 1 and Node 2 both are active , when Listener is on Node1 , application is running fine but when failed to Node 2 , application is not opening giving ASP error.

    What could be the reason?

  • Hi,

    are all accounts existing on both nodes?

    Take a look in your ERRORLOG on each node.

    Andreas

  • What is the error that the application reports?

    Thomas Rushton
    blog: https://thelonedba.wordpress.com

  • Have a look on the error logs before you take any further step.

  • Is the application hosted on machines on different subnets than the AG nodes? I ask because I had an issue in our environment where app servers would lose communication for about 13-15 minutes after a failover. The reason turned out to be a security setting on our data center switches preventing gratuitous ARP packets. As a results when a failover occurred the switches ignored the packets from the AG nodes saying the listener was now associated with a new MAC addresss. This would cause an outage until the ARP table entry natually expired and the switches requested updated information.

    In our situation the only fixes were to either allow gARP packets at the data center switch level or add additional NICs to application servers to get them talking to SQL Server on the same subnet so when a failover occurred their local ARP tables were updated immediately. We went with the latter solution.

    Joie Andrew
    "Since 1982"

Viewing 5 posts - 1 through 4 (of 4 total)

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