Hi,
I've got a situation, where I've got:
On my home site the 2xSRX240H2 in cluster, on the POP1 MX80 and on the POP2 another MX80.
ge-0/0/1 and ge-0/0/2 on both routers are connected as control and fabric, but this does not matter, as I'm having same situation on the cluster with SRX1400 series.
The problems is routing.
I'm getting some routes and default GW from POP1 MX80 (default gateway from POP1 has higher local pref configured) and some routes from POP2 MX80. Scheme is pretty straight-forward, SRX is connected to both MX-s in that way where ge-0/0/12 is used to connect to POP1 and ge-5/0/12 to POP2. The ge-0/0/12 and ge-5/0/12 are not RETH devices. Just looks like this guy's http://forums.juniper.net/t5/SRX-Services-Gateway/Active-Passive-cluster-and-BGP/td-p/242122
So the actual problem, that I'm not able to ping any route received from POP2 from the lo0 interface, while default gw and routes from POP1 are working fine. As soon as I configure some more networks on different interfaces, I can ping them from those networks, but still not able to from lo0 interface, which is used for BGP connections between SRX and MX-s (which means I can't get a working connection from WAN to my SRX if this WAN is located behind routes I'm getting from POP2)
So I see the problem here that the passive SRX in cluster does not have a RE running and packets can't come back from those routes from POP2? Or fab interfaces do not work as supposed to and packets are not being forwarded from one SRX to another? Do I really have to run a full mesh connections (2 cables per POP in RETH devices) to achieve my goal? Or do I have to configure redundancy groups so the ge-5/0/12 would be in active mode?
Need an advice from an experts or guys who had same situations. I've really tried and read a lot, before I've post here...