Redis [BUG] redis-cli add new node hanging forever when a node is in a connected failure state

Describe the bug

redis-cli add-node hangs when there is a node in a connected failure state

To reproduce

The cluster is in the following state: 6d4ac11762296309ceeb6632f978b07d02269e15 10.144.6.13:6379@16379 slave 6f6b775aa7a88c8a8c26e6ab1506c0aacd982529 0 1639739165247 8 connected 000e71bfa6a1f3ad4a230071c5c51fc37ae0c499 10.144.3.13:6379@16379 slave 0042335d7b47fae143906e811919b0539535d288 0 1639739165247 9 connected 5d33620b598abb44058fdf2a1448c237d02989dc 10.144.7.13:6379@16379 master - 0 1639739165248 3 connected 10923-16383 231fb0d1e1398985709d95d97a515f844ee71805 10.144.8.6:6379@16379 myself,slave 5d33620b598abb44058fdf2a1448c237d02989dc 0 1639739165000 3 connected 340667eacba01e3aff4fbd4026863cb40bfc87f4 10.144.27.14:6379@16379 slave 5d33620b598abb44058fdf2a1448c237d02989dc 0 1639739165247 3 connected 0042335d7b47fae143906e811919b0539535d288 10.144.9.14:6379@16379 master - 0 1639739165247 9 connected 5461-10922 20d43990c92f5fa7fcaab0676d62fdfbe00aebfe 10.144.15.3:6379@16379 master,fail - 1639557068707 1639557068607 1 connected 6f6b775aa7a88c8a8c26e6ab1506c0aacd982529 10.144.39.15:6379@16379 master - 0 1639739165248 8 connected 0-5460

10.144.15.3 got disconnected and is now gone.

if you run redis-cli ${REDIS_TLS} ${REDIS_AUTH} --cluster add-node ${NEW_IP}:${NEW_PORT} 10.144.9.14:${REDIS_PORT} --cluster-slave

Adding node 10.144.5.5:6379 to cluster 10.144.9.14:6379 ^C

The command hungs indefinitely.

If I run a CLUSTER MEET on a connected node everything works. If I remove node 20d43990c92f5fa7fcaab0676d62fdfbe00aebfe and retry the command, it works fine.

Expected behavior

A node should be added even if there are disconnected nodes in the list