X-Git-Url: https://git.opendaylight.org/gerrit/gitweb?a=blobdiff_plain;f=test%2Ftools%2Fclustering%2Fcluster-monitor%2FREADME.markdown;fp=test%2Ftools%2Fclustering%2Fcluster-monitor%2FREADME.markdown;h=0000000000000000000000000000000000000000;hb=59e81c38620fa1b61e15771191e35771450b9499;hp=113db9382e268620c960f65b72d4fda6f3aac98b;hpb=072f6e3a8d1bdf8f4c663843589c22d93ba07791;p=integration%2Ftest.git diff --git a/test/tools/clustering/cluster-monitor/README.markdown b/test/tools/clustering/cluster-monitor/README.markdown deleted file mode 100644 index 113db9382e..0000000000 --- a/test/tools/clustering/cluster-monitor/README.markdown +++ /dev/null @@ -1,71 +0,0 @@ -# Cluster Monitor Tool - -This tool provides real-time visualization of the cluster member roles for all -shards in the config datastore. It is useful for understanding cluster behavior -in when controllers are isolated, downed, or rebooted. - -A file named `cluster.json` containing a list of the IP addresses of the -controllers is required. This resides in the same directory as `monitor.py`. -"user" and "pass" are not required for `monitor.py`, but they may be -needed for other apps in this folder. Because this configuration -information unique to your environment, it may be more convenient to -copy the contents of this folder out of git to prevent these settings -from being overwritten by updates. - - -The file should look like this: - -``` - { - "cluster": { - "controllers": [ - "172.17.10.93", - "172.17.10.94", - "172.17.10.95" - ], - "user": "username", - "pass": "password" - } - } -``` - -## Usage: `monitor.py` - -### Starting `monitor.py` - -Before using, start and configure all controllers in the cluster. Use of the -cluster deployment script is recommended. All controllers must initially be -running so the tool can retrieve the controller and shard names. Once -the tool is started and the controller and cluster shard names are retrieved, -controllers can be isolated, downed, rebooted, etc. - -### The `monitor.py` UI - -Controller member names (not host names) are displayed across the top. Shard -names are displayed to the left. - -In the upper left is a heart emoticon "<3" which toggles between yellow and -black backgrounds with each update. If a controller is down, the HTTP timeout -comes in to play and updating becomes much slower. - -The central matrix displays controller roles. When REST queries fail, the -error type is displayed. Leader, Follower, and Candidate roles are color- -coded. - -## Other Scripts - -### `isolate.py` - -Isolates an indicated controller from the cluster. - -### `rejoin.py` - -Rejoins any isolated controllers to the cluster. - -### `timed_isolation.py` - -Isolates an indicated controller for a specified duration - -## Future Enhancements - -Add operational shards.