This is not a bug with the code in particular, but due to lack of forum that I
was able to find I'm hoping someone here might be able to suggest something to
get around this problem.
What steps will reproduce the problem?
1. Perform a failover that includes use of the master_ip_online_change script.
2. master_ip_online_change uses a basic ssh command to bring up the new vip IE:
ssh user@host "sudo /sbin/ifconfig eth0:1 10.10.10.10/16"
What is the expected output? What do you see instead?
When the failover occurs, it takes the ip away from the original host and
populates it on the new host. This means that the IP has not changed so far as
the network landscape is concerned, however when you access that IP the MAC
address that responds is different, which is causing our firewall to panic and
not allow connections to be established. We have tried using /sbin/arping via
ssh from the mha manager, but it hangs, so if we put this in the
master_ip_failover script, the whole script hangs and the failover does not
complete.
Has anyone else run into this problem?
What version of the product are you using? On what operating system?
0.56 on Centos 5
Please provide any additional information below.
Original issue reported on code.google.com by peter.t....@gmail.com on 15 Dec 2014 at 8:49
Original issue reported on code.google.com by
peter.t....@gmail.comon 15 Dec 2014 at 8:49