/
monitors.ad
28 lines (21 loc) · 1.8 KB
/
monitors.ad
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
=== Monitors
ifdef::pro[]
The Monitors screen allows you to view events of system problems recorded by both local and remote nodes.
The list of events can be filtered to limit the number of events displayed, event type, severity level, and node ID.
Filtering by severity level will match the level you choose and any level above it.
Events are listed in descending order by event time, but the order can be changed by clicking column headings.
The remove button will clear the event from the table on the current node. Events are purged automatically each night
using the `purge.retention.minutes` parameter to remove by event time.
image::manage/manage-monitors.png[]
endif::pro[]
When a <<_monitors,Monitor>> is configured, it is run periodically to check the current value of a system metric and compare it to a threshold value.
Different monitor types can check the CPU usage, disk usage, memory usage, batch errors, outstanding batches, unrouted data, and number
of data gaps.
Custom monitor types can be created using <<_extensions,Extensions>> that use the IMonitorType interface.
When the value returned from the check meets or exceeds the threshold value, a <<_monitor_event>> is recorded.
The <<_monitor_event>> table is synchronized on the heartbeat channel, which allows a central server to see events from remote nodes,
but this behavior can be disabled by setting the `monitor.events.capture.enabled` parameter to false.
To be immediately notified of a monitor event, use <<_notifications,Notifications>> to match on the severity level.
Different notification type can send a message by writing to the log or sending an email.
Custom notification types can be created using <<_extensions,Extensions>> that use the INotificationType interface.
In order to send email, the <<_mail_server,Mail Server>> should be configured.