Nodes participating in a cluster are connected with each other through TCP connections which may break up due to network failure. In that case, a client node tries to reestablish the connection. It goes through all the servers in the list once and if none is available, then it sleeps for Retry Interval seconds and then scans the list again. This is done Connection Retries times. In other words, the value of Connection Retries specifies how many times the node should try to restore the connection. Connection Retries can be changed in the following way:
Using NCache Manager
Click on the cache name in Cache Explorer to open cache settings.
Go to Cluster Settings tab.
Set the Connection Retries as per your requirement.
Right click on the cache name in Cache Explorer and select Apply Configuration option.
Manually Editing NCache Configuration
For Community Edition, you can manually edit config.ncconf located at %NCHOME%/config.
Make sure Cache Host Process is stopped before making any configuration changes. If it is not stopped and the changes are made, kill the Cache Host process and then start it.
- Specify connection retries count through config.ncconf by specifying the
connection-retrieskey in the
<cluster-settings ...> <cluster-connection-settings connection-retries="2" .../> </cluster-settings>
- Once changes are made, start the Cache Host process.