[BUG]:master failed to restart if local journal is lost

[shuoranliu]

Describe the bug
In a 3-master-node cluster, if we delete one master pod and its local meta and journal, it failed to restart with the following raft panic. So it will never join the raft group again and catch up with other member's commit index.

SERVER_TYPE: master, ACTION_TYPE: start
POD_IP: 
POD_NAMESPACE: 
POD_CLUSTER_DOMAIN: 
Loading environment variables from /app/curvine/conf/curvine-env.sh
Starting master service...
datetime: 2025-12-25 06:12:55.498, git version: 0e69244, args: ServerArgs {
    service: "master",
    conf: "/app/curvine/conf/curvine-cluster.toml",
}
Applying master hostname from env: 'cv-master-0.cv-master.curvine.svc.cluster.local'
2025-12-25 06:12:55.696 Panic occurred at /root/.cargo/registry/src/mirrors.aliyun.com-0671735e7cc7f5e7/raft-0.7.0/src/raft_log.rs:292: to_commit 1 is out of range [last_index 0], raft_id: 1, backtrace <disabled>

To Reproduce

1. Deploy a 3-master-node cluster.
2. Kill one master pod, and delete its data disk or just delete its local meta and journal.
3. Restart this pod with the same master hostname.

Expected behavior
The master pod restarts successfully, and catch up with the raft commit index.

Screenshots
If applicable, add screenshots to help explain your problem.

OS Version (please complete the following information):

Additional context
Add any other context about the problem here.