Upvotes 0 Share PDF

Error retreiving usage metrics observed in UCP UI


  • The following usage metrics errors can be observed from the UCP UI:

    Error retrieving cpu usage metrics Error retrieving memory usage metrics Error retrieving disk usage metrics

  • The ucp-metrics container does not start and appears to be Exited in docker ps -a output:

    $ docker ps -a --filter name=ucp-metrics CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES bfafbfc679e4 docker/ucp-metrics:2.1.4 "/bin/entrypoint.s..." 4 days ago Exited (1) About an hour ago ucp-metrics

  • Running the docker logs command against the ucp-metrics container reveals the following error:

    $ docker logs -t ucp-metrics 2017-07-10T23:43:59.173335762Z time="2017-07-10T23:43:59Z" level=error msg="Error opening memory series storage: leveldb/storage: corrupted or incomplete meta file" source="main.go:181"


  1. Clear the corrupted metrics database using the commands below. All historical metrics will unfortunately be lost in this procedure:

    # docker run --rm -it -v ucp-metrics-data:/data alpine sh
    # rm -rf /data/* 
  2. Remove the metrics container and trigger the ucp-reconciler to run and reconcile the metrics container again:

    # docker rm ucp-metrics -f
    # docker start ucp-reconcile 
    # docker logs -f ucp-reconcile

Once the reconciliation completes, metrics should be functional on this UCP node once again.