Master reboot is detected by sentinel but it doesn't trigger failover.
[7678] 24 Sep 10:53:16.164 * +reboot master redis1 192.168.250.128 2148
Keeping old master will create possibility for data loss as slave sync is usually faster that persistence. Using AOF and syncing to disc with every write is the only case when master should have newer data.
Master reboot is detected by sentinel but it doesn't trigger failover.
[7678] 24 Sep 10:53:16.164 * +reboot master redis1 192.168.250.128 2148
Keeping old master will create possibility for data loss as slave sync is usually faster that persistence. Using AOF and syncing to disc with every write is the only case when master should have newer data.