[OpenSIPS-Users] one node in 3 node cluster thinks all gateways are down

[Kingsley Tart]

Hi,

I have a 3 node cluster of OpenSIPS 3.1.6 with VIPs maintained by
keepalived. Each has its own preferred VIP but will take on others if
other nodes become unavailable. Primarily, node A has 'vip-a', node B
has 'vip-b', and node C has 'vip-c'.

Node A is normally the one doing gateway probing:

modparam("drouting", "cluster_sharing_tag", "vip-a")

Right now, node A still has vip-a and is doing the gateway probing.
Node A and node B both consider all gateways active, but node C
considers them all inactive.

Node C is still receiving messages from the cluster, so I don't
understand why this is, given that it's only node A that's doing the
probing (I've verified this with ngrep) - only node A is originating
OPTIONS messages).

Any clues?

Cheers,
Kingsley.