无法连接Cassandra服务器，服务运行正常但无法恢复连接求助

Hey there, let's work through this Cassandra problem you're facing. That active (exited) status from sudo service cassandra status is a bit misleading—it means the legacy SysV init script ran and exited successfully, but it doesn't confirm the Cassandra daemon is actually running in the background and ready to accept connections. Here's how to dig deeper:

Step 1: Verify Cassandra is actually running #

First, check if the Cassandra Java process is active on your system:

ps aux | grep cassandra

Look for a process owned by the cassandra user with a long Java command line. If you don't see this, the service didn't start properly—skip straight to checking logs.

Step 2: Check Cassandra's system logs #

The most valuable clues will be in Cassandra's system log. By default, this lives at /var/log/cassandra/system.log. Use this command to view the latest entries in real time:

tail -f /var/log/cassandra/system.log

Keep an eye out for error messages like:

• Port conflicts (e.g., another service using 9042, Cassandra's default CQL port)
• Heap memory issues (not enough allocated RAM for the daemon)
• Typos or invalid settings in cassandra.yaml
• File permission problems on data or log directories

Step 3: Validate key configuration settings #

Open your cassandra.yaml file (typically at /etc/cassandra/cassandra.yaml) and double-check these critical values:

• listen_address: Should be the IP address your client (or other nodes) can reach. For local connections, localhost works; for remote access, use the server's public/private IP.
• rpc_address: This handles client CQL connections—ensure it's accessible to your machine.
• native_transport_port: Default is 9042. Confirm no other service is using this port with:

netstat -tulpn | grep 9042

Step 4: Check firewall rules #

If Cassandra is running but you still can't connect, your firewall might be blocking the CQL port. Verify your firewall status:

• For UFW:

sudo ufw status

If there's no rule allowing traffic on 9042, add it with sudo ufw allow 9042.

• For iptables:

sudo iptables -L

Look for a rule that accepts incoming traffic on port 9042.

Step 5: Restart Cassandra with systemd commands #

Instead of using the legacy service command, use systemd directly for more accurate status reporting:

sudo systemctl restart cassandra
sudo systemctl status cassandra

A healthy Cassandra daemon should show active (running) here, which is far more reliable than the ambiguous active (exited) from the SysV script.

Step 6: Check the PID file #

Cassandra creates a PID file at /var/run/cassandra/cassandra.pid by default. Check if the PID in this file matches a running process:

cat /var/run/cassandra/cassandra.pid
ps aux | grep <PID_FROM_FILE>

If the PID doesn't correspond to a running Cassandra process, the daemon crashed after the init script exited.

Start with these steps, and the system log will almost certainly point you to the root cause—whether it's a config typo, resource limit, or network issue.

内容的提问来源于stack exchange，提问作者Bmaed Riasbet