Messaging Gateway

  • 1.  DNS Round Robin Failover Not Working

    Posted Jul 06, 2009 05:15 AM

    Hi,
    i have configured 2 BrightMail Gateway appliances as below:

    - Appliance 1 - Control Center + Scanner for inbound message
    - Appliance 2 - Scanner for Inbound + Outbound message

    I configured DNS Round Robin to load balance the inbound message to the 2 appliances with DNS configuration below:

    IN     MX     10     mail.abc.com
    IN     MX     10     mail2.abc.com

    mail     IN     A     68.25.218.11 (Appliance 1)
    mail2   IN     A     68.26.218.12 (Appliance 2)

    In the Firewall configuration, the mapping is as below:

    68.25.218.11 mapped to 10.10.10.1
    68.26.218.12 mapped to 10.10.10.2

    The load balance is running fine which first 10 inbound messages will go Appliance 1 and subsequent inbound messages (11 to 20) will go to Appliance 2. when i unplugged network interface of 10.10.10.1 on Appliance 1 to simulate failure on Appliance, the inbound messages did not go to Appliance 2 which mean no inbound message coming into the network.

    Anyone has experienced this before? appreciate your advice.

    Thanks
    Yap
     



  • 2.  RE: DNS Round Robin Failover Not Working

    Posted Jul 06, 2009 06:15 AM
    Hello Yap Kim Siong,

    If you are having 2 scanners & one control center then both the scanner will communicate with that control center. If you unplugged the server which is having controlcenter then the mail genuine emails will be delivered normally to the end users, but the mail with action Hold message in Spam Quarantine will not be able to view & will be drop as it is not able to communicate wi the control center.

    Kindly check by sending mails from external domain that what error you get if your controlcenter is down.

    Regards,
    Laeek


  • 3.  RE: DNS Round Robin Failover Not Working

    Posted Jul 06, 2009 07:53 AM
    Hi Laeek,
    Thanks for your kind feedback, i have scheduled another failover test tomorrow. The test scenario will be same as the previous attempt except i requested the DNS administrator to refresh the DNS cache or restart the DNS service. The action plan is as below:

    - unplug network interface of inbound message on appliance functioned as Control Center + Scanner for inbound message
    - Refresh DNS cache on DNS server
    - monitor inbound message but establishing SSH session 2nd Appliance.

    is there other thing that i need to check or miss out?

    Thank you very much in advance

    cheers,
    yap