Forum Discussion
CirrusF5 BGP Peering in Active /Standby Cluster
Hi I was able to advertise routes by declaring a floating Ip for the bgp peering subnet self ips and attaching it to default traffic group , rest F5 takes care of advertising routes only via floating ip on both active and standby
MVPlet say you dont use f5 but a pair cisco routers as the pool member's default gw,
both cisco routers will advertise pool member's subnet isnt it?
f5 zebos-based tmrouted, including on standby node, does the same.
https://my.f5.com/manage/s/article/K10168
so for example, when admin wants to ssh the pool member, it's ok if the ip traffic goes through standby f5.
just like when you use non f5, if you dont want users to access pool member app ports but must be through f5 vserver,
then you need to configure network firewall or f5 acl accordingly.
rhi is intended to advertise f5 vip, which surely wont be active in standby node.
Shripatyhi , yes that is correct , the f5 is acting as two arm here the external vlan will be used for vip and for that i am able to inject RHI feature successfully whereas the Internal VLAN being used for Pool Member subnets has its own self ip and floating ip which will be used by Webservers as their default gateway to achieve Direct Server Return policy. The TMM route is being advertised successfully for External VLAN but the problem I m facing is how to control the route advertisement for Internal VLAN since the redistribute connected command does advertises both the external connected subnet and as well internal connected subnet from active and standby
