chore: modify replicasync to handle NATS explicitly (#26666)

relates to GRU-69

Modifies replicasync to handle discovering NATS enabled primary replicas explicitly, and passing that info to the NATS Pubsub.

This PR adds a new deployment value to explicitly represent the host or IP that the replica can be reached on. It isn't wired up to the CLI, but piggybacks on the DERP config for now.

We learn the NATS port directly from NATS at runtime, and propagate it thru replicasync to learn all peers for clustering.
This commit is contained in:
Spike Curtis
2026-06-26 08:36:32 -04:00
committed by GitHub
parent 9f211ce5ae
commit 98e1ce133c
16 changed files with 219 additions and 132 deletions
+4 -1
View File
@@ -678,7 +678,8 @@ func New(ctx context.Context, options *Options) (_ *API, err error) {
}
// We always want to run the replica manager even if we don't have DERP
// enabled, since it's used to detect other coder servers for licensing.
// enabled, since it's used to detect other coder servers for licensing,
// and NATS clustering for HA pubsub.
api.replicaManager, err = replicasync.New(ctx, options.Logger, options.Database, options.ReplicaSyncPubsub, &replicasync.Options{
ID: api.AGPL.ID,
RelayAddress: options.DERPServerRelayAddress,
@@ -686,6 +687,7 @@ func New(ctx context.Context, options *Options) (_ *API, err error) {
RegionID: int32(options.DERPServerRegionID),
TLSConfig: meshTLSConfig,
UpdateInterval: options.ReplicaSyncUpdateInterval,
ClusterHost: options.ClusterHost,
})
if err != nil {
return nil, xerrors.Errorf("initialize replica: %w", err)
@@ -788,6 +790,7 @@ type Options struct {
// Used for high availability.
ReplicaSyncUpdateInterval time.Duration
ReplicaErrorGracePeriod time.Duration
ClusterHost string // IP or hostname to reach this specific replica
DERPServerRelayAddress string
DERPServerRegionID int