I've got 2 CSS 11506 boxes configured using box to box failover. Failing the master CSS box itself (powering down) causes the backup CSS to become master and all is well.
However when the switch, which the CSS is connected to, fails the CSS didn't fail over so I added the redundancy-phy to both the interfaces connected to the switch and failed the switch again. At this point a "show redundancy" shows the master becoming backup but between 3 and 5 seconds later it re-assumes master status and keeps flipping every 60 - 90 seconds
I also tried a service with a type of redundancy-up and again the same symptoms - fails over but assumes master again within 3--5 seconds.
Any help gratefully received!