• Cause: The versions earlier than ZStack 3.5.0 use Prometheus 1.8.2, which might cause the monitoring data to occupy the root space.
    Note: ZStack 3.5.0 and later versions use Prometheus 2.9.2, which allow you to set the retention policy. (By default, 50 GB monitoring data can be saved.)
  • Solution: If the monitoring data is no longer needed, permanently delete the Prometheus monitoring data to release the space.
    • In a single management node environment, run the following commands in sequence to delete the Prometheus monitoring data:
      [root@hostname ~]# du -sh /var/lib/zstack/prometheus/data/    //View the monitoring data size.
[root@hostname ~]# zstack-ctl stop 
[root@hostname ~]# kill -9 `ps -ef |grep prometheus|awk '{print $2}'` 
[root@hostname ~]# rm -rf /var/lib/zstack/prometheus/data
[root@hostname ~]# zstack-ctl start
    • In a dual-MN environment, run the following commands in each management node to delete the Prometheus monitoring data:
      

      [root@mn-1 ~]# du -sh /var/lib/zstack/prometheus/data/    //View the monitoring data size.
[root@mn-1 ~]# zsha2 stop-node
[root@mn-1 ~]# kill -9 `ps -ef |grep prometheus|awk '{print $2}'
[root@mn-1 ~]# rm -rf /var/lib/zstack/prometheus/data
[root@mn-1 ~]# zsha2 start-node