Description:
(Please note: English is not my native language, sorry for any mistakes.)

In a long-lived PHP process (e.g., Laravel Queue Worker), RedisCluster fails to update its slot map when a previously promoted Slave becomes a Master again. This leads to a persistent error state in the worker even after the cluster is healthy.

Steps to Reproduce:
State 1: Node-A (Master), Node-B (Slave). Worker: OK.
State 2: Node-A fails. Node-B is promoted to Master. Worker: OK.
State 3: Node-A returns as a Slave of Node-B. Worker: OK.
State 4: Node-B fails. Node-A is promoted back to Master.
Actual Result: Worker throws RedisClusterException: Redis Cluster's master data is incomplete. Cluster is in invalid state.
State 5: Node-A is Master, Node-B is Slave (Cluster is fully recovered and healthy).

Actual Result: The worker process continues to throw the exception and cannot recover the keyspace map.

Technical Details:
The exception is triggered in cluster_library.c during the slot update process:

redisClusterNode *prev_master = SLOT(c,c->redir_slot);
if (NULL == prev_master->slaves) {
    CLUSTER_THROW_EXCEPTION("Redis Cluster's master data is incomplete. Cluster is in invalid state.", 0);
    return FAILURE;
}

In State 4 and State 5, the prev_master->slaves pointer remains NULL in the worker's memory, causing the remap process to abort before the new topology can be applied.