Mesos Agent持续处于Deactivated状态问题求助

Hey there, let's work through why your Mesos Slave is stuck in the Deactivated state. Based on the cluster setup you described (VMware WorkStation with heron01 as Master/Zookeeper and heron02 as Slave), here are the most common troubleshooting steps to start with:

1. Verify Network Connectivity Between Nodes #

First, make sure the two VMs can communicate properly:

• On heron02, run ping 192.168.201.131 to check basic connectivity. If this fails, you'll need to fix VMware's network settings first (confirm they're on the same subnet, check firewall rules, etc.).
• Test if Mesos ports are reachable:
  • Master's default HTTP port is 5050; run telnet 192.168.201.131 5050 or nc -zv 192.168.201.131 5050 from heron02 to confirm.
  • Slave needs to communicate with Master on port 5051 too, so test that port as well.

2. Check Zookeeper Configuration Consistency #

Mesos relies on Zookeeper for service discovery—any mismatch here will break slave registration:

• Open mesos-master-env.sh on heron01 and look for the MESOS_ZK parameter. It should look something like zk://192.168.201.131:2181/mesos.
• Now check mesos-slave-env.sh and mesos-agent-env.sh on heron02: their MESOS_ZK value must be exactly the same as the master's. Even a typo in the IP or port will prevent the slave from finding the master.

3. Validate Slave Resource Configuration #

Sometimes slaves get deactivated if their declared resources don't match the actual system capacity:

• Check the MESOS_CPUS, MESOS_MEM, MESOS_DISK parameters in your slave config files. Make sure these values don't exceed the actual CPU cores, memory, or disk space available on heron02.
• Run free -h and lscpu on heron02 to verify the real system resources.

4. Ensure Correct File Permissions #

Mesos processes (usually running as the mesos user) need read access to their config files:

• Run these commands on each node to check permissions:

ls -l /etc/mesos/mesos-master-env.sh  # On heron01
ls -l /etc/mesos/mesos-slave-env.sh   # On heron02
ls -l /etc/mesos/mesos-agent-env.sh   # On heron02

If the owner isn't mesos:mesos, fix it with chown mesos:mesos /path/to/config/file.

5. Analyze Logs (Critical Next Step) #

Since you mentioned the Master error logs and Slave warning logs are pending, once you have those, look for these key messages:

• Master logs: Search for lines like Failed to register slave or errors related to authentication/authorization if you have that enabled.
• Slave logs: Look for warnings like Could not connect to master or Failed to connect to Zookeeper—these will point directly to connectivity or config issues.

Once you run through these checks and share the full log snippets or relevant config parts, we can narrow down the exact cause.

内容的提问来源于stack exchange，提问作者Yitian Zhang