Monitoring system performance while it is in operation mode is essential part of making system improvement decisions. Similarly, monitoring cache cluster is important in controlling cache behaviors in order to get maximum benefit from it. NCache provides various options to monitor "NCache Cache Cluster Health" and "Performance Statistics". This document will explain these options and how these can be used for cache cluster health and performance monitoring.
NCache logs all cache level errors/events in "NCache Log Files" and "Windows Event Logs" by default. NCache cache logs show more detailed information and are generally helpful for debugging purposes. NCache cluster events on the other hand can be monitored regularly to get notified for any cluster level issues.
NCache also provides extensive set of performance counters for cache server and client performance statistics. These are very helpful in regular NCache performance monitoring as well as for debugging performance issues.
System administrators can use NCache monitoring capabilities or any third party tools using these cluster events and performance counters and can take necessary actions accordingly. NCache provides various ways to monitor you cache cluster.
Here are details on various options available for NCache monitoring:
NCache provides cache cluster monitoring with the help of NCache email alert system. These email alerts can be configured and System Administrators can be notified via email for any planned or unplanned cache cluster events such as if cache stops, cache starts, a node leaves or joins a cache etc.
NCache email alerts can be configured from Alerts tab on your cache as follow:
A user can provide their email server details as follows along with recipients email addresses and NCache would send email notifications to all recipients for selected cache events.
Select the events for which the notification is required, listed on the tab. Here are details on events for which an email alert can be enabled:
A system admin can take necessary actions if he finds an email for any of above events and if it is not found to be a planned event.
All cache cluster level events are logged in windows event log by default. You can use NCache email alerts or any 3rd party tool that uses these NCache cluster events to get notified accordingly. It is always a good practice to verify cache cluster health as soon as you see such an event logged in windows event logs. Here are details on important NCache event IDs:
NCache Monitor is a diagnostic tool to analyze performance, health and events of a cache cluster both on a local and remote machine. Cache administrators can monitor cache events, view detailed health information of all nodes in a cluster and monitor performance of a cluster using real time statistics. NCache Monitor can run as a separate application and can monitor any cache from any box in your environment.
NCache Monitor tool can be launched in two ways:
Following are some of the main operations you can perform using NCache Monitor:
Toolbox contains different counters for monitoring performance, health and events of a cache cluster. Widgets are categorized in two groups that are:
1. NCache Server: It contains pre-define NCache server widgets
2. NCache Client: It contains pre-define NCache client widgets
Server Dashboard is one of the default dashboards which appear automatically after the selection of a cluster. You can also retrieve this dashboard at any time from View option as follow:
Please refer to NCache Help documentation for more details on available monitoring options in default server dashboard
Client Dashboard is also one of the default dashboards that can be retrieved any time using View option. Client dashboard contains following graphs:
You can view the client Dashboard from the following option:
This will open another window that can be used to visualize client statistics for monitoring.
Please refer to NCache online help for more details on available monitoring options in default client dashboard.
After selection of a cluster in NCache Monitor, you can create your own dashboards according to your needs and can visualize NCache servers as well as client counters accordingly. To create a new custom dashboard, follow these steps:
On finishing the above steps following screen will be displayed having three rows and a single column as specified above
You can create more than one dashboard for a selected cache cluster
All the open dashboards created against cluster will be automatically saved on closing NCache Monitor, and will be retrieved whenever that cluster is selected for monitoring
The closed dashboards are not saved for a selected cluster