Fixed head offset in follower after a snapshot load

[merlimat]

The head snapshot in follower was not updated after loading a snapshot. This can make it look like the follower is not caught up with the leader, when in fact it is.

Dec 27 22:43:34.180 |INF| Successfully moved ensemble to a new term component='shard-controller' followers=[{ server-address={ public='oxia-2.oxia-svc.oxia.svc.cluster.local:6648' internal='oxia-2.oxia-sv
c:6649' } entry-id={ term=-1 offset=-1 } },{ server-address={ public='oxia-1.oxia-svc.oxia.svc.cluster.local:6648' internal='oxia-1.oxia-svc:6649' } entry-id={ term=-1 offset=-1 } }] namespace='default' n
ew-leader={ public='oxia-5.oxia-svc.oxia.svc.cluster.local:6648' internal='oxia-5.oxia-svc:6649' } shard=1 term=19
Dec 27 22:43:34.229 |INF| Successfully deleted shard component='shard-controller' namespace='default' server={ public='oxia-4.oxia-svc.oxia.svc.cluster.local:6648' internal='oxia-4.oxia-svc:6649' } shard=
1
Dec 27 22:43:34.233 |INF| Elected new leader component='shard-controller' leader={ public='oxia-5.oxia-svc.oxia.svc.cluster.local:6648' internal='oxia-5.oxia-svc:6649' } namespace='default' shard=1 term=1
9
Dec 27 22:43:34.241 |INF| Follower is *not* caught-up yet with the leader component='shard-controller' follower-head-offset=-1 leader-head-offset=27556 namespace='default' server={ public='oxia-2.oxia-svc
.oxia.svc.cluster.local:6648' internal='oxia-2.oxia-svc:6649' } shard=1
Dec 27 22:43:34.324 |INF| Follower is *not* caught-up yet with the leader component='shard-controller' follower-head-offset=-1 leader-head-offset=27556 namespace='default' server={ public='oxia-2.oxia-svc
.oxia.svc.cluster.local:6648' internal='oxia-2.oxia-svc:6649' } shard=1
Dec 27 22:43:34.406 |INF| Follower is *not* caught-up yet with the leader component='shard-controller' follower-head-offset=-1 leader-head-offset=27556 namespace='default' server={ public='oxia-2.oxia-svc
.oxia.svc.cluster.local:6648' internal='oxia-2.oxia-svc:6649' } shard=1
Dec 27 22:43:34.672 |INF| Follower is *not* caught-up yet with the leader component='shard-controller' follower-head-offset=-1 leader-head-offset=27556 namespace='default' server={ public='oxia-2.oxia-svc
.oxia.svc.cluster.local:6648' internal='oxia-2.oxia-svc:6649' } shard=1
Dec 27 22:43:34.922 |INF| Follower is *not* caught-up yet with the leader component='shard-controller' follower-head-offset=-1 leader-head-offset=27556 namespace='default' server={ public='oxia-2.oxia-svc
.oxia.svc.cluster.local:6648' internal='oxia-2.oxia-svc:6649' } shard=1
Dec 27 22:43:35.621 |INF| Follower is *not* caught-up yet with the leader component='shard-controller' follower-head-offset=-1 leader-head-offset=27556 namespace='default' server={ public='oxia-2.oxia-svc
.oxia.svc.cluster.local:6648' internal='oxia-2.oxia-svc:6649' } shard=1
Dec 27 22:43:36.113 |INF| Follower is *not* caught-up yet with the leader component='shard-controller' follower-head-offset=-1 leader-head-offset=27556 namespace='default' server={ public='oxia-2.oxia-svc
.oxia.svc.cluster.local:6648' internal='oxia-2.oxia-svc:6649' } shard=1

The node had indeed correctly applied the snapshot:

Dec 27 22:38:34.193 |INF| Received SendSnapshot request component='internal-rpc-server' namespace='default' peer='10.244.0.130:50430' shard=1
Dec 27 22:38:34.193 |INF| Closing leader controller component='leader-controller' namespace='default' shard=1 term=18
Dec 27 22:38:34.194 |INF| [JOB 1] WAL file /data/db/default/shard-1/000002.log with log number 000002 stopped reading at offset: 0; replayed 0 keys in 0 batches component='pebble' shard=1
Dec 27 22:38:34.196 |INF| Created follower commit-offset=-1 component='follower-controller' head-offset=-1 namespace='default' shard=1 term=18
Dec 27 22:38:34.208 |INF| Successfully applied snapshot commit-offset=27556 component='follower-controller' namespace='default' shard=1 snapshot-size=449137 term=18
Dec 27 22:38:34.220 |INF| Received Replicate request component='internal-rpc-server' namespace='default' peer='10.244.0.130:50430' shard=1