t8y2/dbx · error

send ZooKeeper connect request: %w

Error message

send ZooKeeper connect request: %w

What it means

newProtocolZooKeeperClient builds the ZooKeeper connect request and writes it as a single frame with writeFrame; if the write fails (connection broken, closed, timed out) the error is wrapped as 'send ZooKeeper connect request'. This happens after the TCP/TLS connection succeeded but before any server response, indicating a transport-level problem during handshake.

Source

Thrown at agents/drivers/argo-go/zookeeper_protocol.go:156

	closed     bool
}

func newProtocolZooKeeperClient(connection net.Conn, timeout time.Duration) (*protocolZooKeeperClient, error) {
	if connection == nil {
		return nil, errors.New("ZooKeeper connection is nil")
	}
	if timeout <= 0 {
		timeout = defaultConnectTimeout
	}
	client := &protocolZooKeeperClient{connection: connection, timeout: timeout}
	request := &zooKeeperEncoder{}
	request.int32(zooKeeperProtocolVersion)
	request.int64(0)
	request.int32(zooKeeperTimeoutMillis(timeout))
	request.int64(0)
	request.bytes(make([]byte, 16))
	if err := client.writeFrame(request.data()); err != nil {
		return nil, fmt.Errorf("send ZooKeeper connect request: %w", err)
	}
	response, err := client.readFrame()
	if err != nil {
		return nil, fmt.Errorf("read ZooKeeper connect response: %w", err)
	}
	decoder := newZooKeeperDecoder(response)
	if _, err := decoder.int32(); err != nil {
		return nil, fmt.Errorf("decode ZooKeeper protocol version: %w", err)
	}
	if _, err := decoder.int32(); err != nil {
		return nil, fmt.Errorf("decode ZooKeeper session timeout: %w", err)
	}
	sessionID, err := decoder.int64()
	if err != nil {
		return nil, fmt.Errorf("decode ZooKeeper session ID: %w", err)
	}
	if _, err := decoder.bytes(); err != nil {
		return nil, fmt.Errorf("decode ZooKeeper session password: %w", err)

View on GitHub (pinned to c0390bff16)

Solutions

1. Check connectivity to the ZooKeeper host:port and any load balancer/idle-timeout in front of it
2. Ensure the ZooKeeper server is up and accepting connections (server log, 4lw 'ruok' command)
3. Increase the dial/session timeout if writes time out on slow links
4. Retry the connection — connectKerberosZooKeeper will fail over to the next ensemble member

conn, err := net.DialTimeout("tcp", address, 5*time.Second)
if err != nil { return err }
conn.Close() // reachability pre-check before the protocol handshake

client, err := newProtocolZooKeeperClient(conn, timeout)
if err != nil {
    if errors.Is(err, syscall.ECONNRESET) || isTimeout(err) {
        return retryWithBackoff(3, func() error { return dialAndConnect(address) })
    }
    return err
}

When it happens

Trigger: Calling connectKerberosZooKeeper (which invokes newProtocolZooKeeperClient) where the underlying net.Conn write of the connect frame fails: server closed the connection immediately after accept, network timeout mid-write, or TLS handshake accepted but peer reset the stream.

Common situations: ZooKeeper listener closed by a load balancer idle timeout; packet loss on the way to the ensemble; server crash between TCP accept and protocol read; Docker/K8s sidecar terminating the connection before the first frame.

Related errors

• ZooKeeper SASL response is truncated
• read ZooKeeper connect response: %w
• ZooKeeper connection timed out before a session was establis
• ZooKeeper session expired during connection
• ZooKeeper sent an unexpected token after GSSAPI completion

AI-assisted analysis of t8y2/dbx@c0390bff16 (2026-09-05). Data as JSON: /api/errors/22a6c7f5da5029e5. Report an issue: GitHub.