multi az support

[wallies]

No description provided.

[wallies]

this should solve #66

[tayzlor]

whats the significance of adding the retry join?

[wallies]

consul was failing to elect a leader and then failing to show up as services in the UI. Adding this fixed it.
-retry-join - Similar to -join but allows retrying a join if the first attempt fails. This is useful for cases where we know the address will become available eventually.

[pmbauer]

fwiw, I've seen consul consistently fail to cluster on bootstrap/apollo-launch.sh and every consul agent ends up with a cluster of one member, itself. We worked around it by defining consul_join but this is more flexible.

[tayzlor]

@pmbauer I've seen that too occassionally - I reported it over in #198. Looks like the consul related fixes here might address that too

[tayzlor]

can we separate the consul retry join to a separate PR as that may resolve issues over in #198 and to avoid blocking on waiting for multi-az

[tayzlor]

Initial overview of the code looks ok. Can we also update the architecture diagram on the frontpage with this change?

To do that you need to provision the cluster as normal, then login to visualops then there is an option to import from AWS (and you can import the running instances) which should create a diagram which you can view/edit

[wallies]

couldnt add variables, as cant have dots in bash variables
export TF_VAR_zones.zone-0=${TF_VAR_zones.zone-0:-eu-west-1a}
export TF_VAR_zones.zone-1=${TF_VAR_zones.zone-1:-eu-west-1b}
export TF_VAR_zones.zone-2=${TF_VAR_zones.zone-2:-eu-west-1c}

[enxebre]

Closing in favour of #385