Notifications and Alerts in Performance Monitoring Toolset
Performance Monitoring Toolset provides end to end monitoring of all events happening inside ColdFusion Server.
There are several ways Performance Monitoring Toolset displays metrics and aggregation over it and it helps user in getting complete picture of how healthy ColdFusion server is.
Out of all these events being monitored by Performance Monitoring Toolset, there are some which need immediate attention of user.
User might be genuinely (or out of curiosity) interested in knowing when these events are happening and how often these events are happening.
That brings us to Alerts/Notification feature of Performance Monitoring Toolset.
How these alerts are raised?
There are several alerts provided with Performance Monitoring Toolset and by default they are disabled.
User can enable and change configuration with relevant values.
Performance Monitoring Toolset keeps evaluating these alerts at every 5 second of interval and if it matches the criterion defined in Alert configuration, then alert is raised and displayed in Performance Monitoring Toolset UI as shown below
Global Filter and Alert configuration:
Alerts can be configured at global filter level which is either at Cluster, Group or CF Server level.
Changing alert configuration at Cluster level will be applied to all CF Servers of that cluster.
Importance of auto-triggered actions :
This is something which is very important and convenient also from usability perspective as it proactively informs user about events even without logging into Performance Monitoring Toolset. It saves user from going through all the hassles of logging into intranet and Performance Monitoring Toolset.
Performance Monitoring Toolset can automatically execute actions when any alert is raised.
Some Generic actions like Sending mail to configured email id or posting to Slack are applicable to all types of alerts and some actions are event specific.
Example of event specific actions is taking heap dump automatically when high heap memory usage alert is raised. In other words, if JVM Heap Memory goes beyond the configured limit then it can be configured to take heap dump automatically.
Based on Customer reported issues related to performance, It has been observed that Heap Dump for reported issue is not available. Sometimes it is difficult to simulate the issue, and lack of heap dump increases the issue resolution time.
In such scenarios, auto-trigger of heap dump can be enabled and it will make sure that heap dump is taken when heap memory usage goes up, and later this heap dump can be analyzed to find the root cause. Heap Dump Analyzer like MAT can be used to find out which section of code is using most of the memory.
User can also write cfc functions and refer to this cfc from alert configuration.