Cirrostratusi want to deploy F5s in different Zones in GCP . reading the below Document . i want to deploy 2 F5s in 2 Zones .
https://clouddocs.f5.com/cloud/public/v1/google/Google_configsync.html
both F5s are behind the Google loadbalancer.
Question - how will config Sync work . meaning will both F5s have the same virtual server ip address ?
where can i download a template to configure the 2 F5s in Config Sync deployment.
Cirrostratusthanks for the reply - the github configuration - and deployment of F5s is extremely complex to figure out . if the only advantage of have a config sync between the 2 F5s is to have "Some components" in sync then it would be easier to do it with CICD -
woudd both F5s have the same IP for a VP?
Nacreousyou need to use dynamic routing based RHI if you want VIP sharing with the f5 entities doesnt have self and floating ip in same subnet
CirrostratusHi All - the problem is -
what would the 2 F5s be Syncing
the Virtual IPS are Different - the pool members are different.
network configuration is different . the only thing common is ldap authentication - dns and nto settings.
what is the advantage of having device sync for that ?
It appears that the only documentation we have is for a configsync in an active/active dual zone environment:
http://clouddocs.f5.com/cloud/public/v1/google/Google_configsync.html
https://clouddocs.f5.com/cloud/public/v1/google_index.html
for
refer below link:
https://clouddocs.f5.com/training/community/adc/html/class2/module1/lab1.html
HTH
🙏
Nacreousthe config sync is like config repository which means f5 devices in the Device Group will store shared data replica (e.g. floating ip, virtual server config but not self (static) ip)
so for example if a device is corrupt and you deply new device, then the new device can simply get the shared config by joining the device group.
vip can be shared (in active-standby mode) only if the devices can have Self IP and Floating IP in Traffic Group (basically same IP subnet; in addition to joining same Device Group).
there is workaround to allow separate traffic group to use same VIP using OSPF/BGP and LTM VS RHI method but i think it's not necessary in your case because the upfront GCP LB will load balance both F5 LBs.
https://techdocs.f5.com/en-us/bigip-14-1-0/big-ip-tmos-routing-administration-14-1-0/dynamic-routing.html#GUID-7CDA4A41-3148-4D5F-B031-45E14190398E
Cirrostratusdoes this mean i need to apply workarounds to make a Active active config work? i am looking for a step by step config for GCP Active Active serup.
