fix: improved socket error logging for connection diagnostics

[vyavdoshenko]

Fixes: #5041

This implementation enhances error diagnostics when replica connections fail by implementing a utility function that extracts and formats detailed TCP socket information from /proc/net/tcp and /proc/net/tcp6 on Linux systems.
The function converts raw socket data into a human-readable format with proper TCP state names and formatted IP addresses. It's integrated with Replica, Connection, and Migration components to provide better context during connection failures.

[romange]

Have you tried killing master/replica in the middle of replication and see what is printed?

[romange]

before trying tcp6, you can get socket family:

int get_socket_family(int fd) {
    struct sockaddr_storage ss;
    socklen_t len = sizeof(ss);

    if (getsockname(fd, (struct sockaddr *)&ss, &len) == -1) {        
        return -1; // Indicate an error
    }

    return ss.ss_family;
}

[vyavdoshenko]

fixed

[romange]

have you actually seen that this code helps?
we do not rely on errno, because iouring does not use it.

[vyavdoshenko]

removed

[vyavdoshenko]

Have you tried killing master/replica in the middle of replication and see what is printed?

Restarting master:

I20250508 15:07:13.919483 14049 replica.cc:705] Transitioned into stable sync
W20250508 15:08:34.295888 14056 common.cc:413] ReportError: Software caused connection abort
E20250508 15:08:34.296075 14056 replica.cc:693] DflyStream error in phase STABLE_SYNC with 192.168.100.3:6379, error: Software caused connection abort, socket state: State: CLOSE_WAIT, Local: 192.168.100.2:55124, Remote: 192.168.100.3:6379, Inode: 233203
W20250508 15:08:34.317605 14056 protocol_client.cc:238] Socket error: Software caused connection abort in 192.168.100.3:6379, socket info: State: CLOSE_WAIT, Local: 192.168.100.2:55124, Remote: 192.168.100.3:6379, Inode: 233203
I20250508 15:08:34.335923 14049 replica.cc:729] Exit stable sync
W20250508 15:08:34.335968 14049 replica.cc:271] Error stable sync with 192.168.100.3:6379 (phase: TCP_CONNECTING): system:103 Software caused connection abort, socket state: socket not found in /proc/net/tcp or /proc/net/tcp6
E20250508 15:08:34.844139 14049 protocol_client.cc:194] Error while calling sock_->Connect(server_context_.endpoint): Connection refused
W20250508 15:08:34.844228 14049 replica.cc:219] Error connecting to 192.168.100.3:6379 (phase: TCP_CONNECTING): system:111, reason: Connection refused
E20250508 15:08:35.345132 14049 protocol_client.cc:194] Error while calling sock_->Connect(server_context_.endpoint): Connection refused
W20250508 15:08:35.345216 14049 replica.cc:219] Error connecting to 192.168.100.3:6379 (phase: TCP_CONNECTING): system:111, reason: Connection refused
I20250508 15:08:35.854527 14049 replica.cc:587] Started full sync with 192.168.100.3:6379
I20250508 15:08:35.855365 14049 replica.cc:607] full sync finished in 8 ms
I20250508 15:08:35.855481 14049 replica.cc:705] Transitioned into stable sync

Restarting replica:

I20250508 15:08:35.846104  1447 dflycmd.cc:696] Registered replica 192.168.100.2:6379
I20250508 15:08:35.854362  1447 dflycmd.cc:393] Started sync with replica 192.168.100.2:6379
I20250508 15:08:35.855278  1447 dflycmd.cc:433] Transitioned into stable sync with replica 192.168.100.2:6379
I20250508 15:09:22.736678  1447 dflycmd.cc:127] Disconnecting from replica 192.168.100.2:6379
W20250508 15:09:22.736735  1447 common.cc:413] ReportError: Operation canceled: ExecutionState cancelled
I20250508 15:09:22.737004  1447 dflycmd.cc:686] Replication error: Operation canceled: ExecutionState cancelled
I20250508 15:09:35.609882  1450 dflycmd.cc:696] Registered replica 192.168.100.2:6379
I20250508 15:09:35.634145  1450 dflycmd.cc:393] Started sync with replica 192.168.100.2:6379
I20250508 15:09:36.694377  1450 dflycmd.cc:433] Transitioned into stable sync with replica 192.168.100.2:6379