The System Events screen shows recent logged events that occur across all the nodes of a distributed AIE system.
You can acknowledge each event, and augment the log record with a comment.
Note that events such as an optimize or an *index delete *execute once per partition across all nodes of the project. As such, they generate multiple event messages.
View incoming links.
Viewing System Events
To view System Events, select Diagnostics -> System Events in the AIE Administrator. The System Events screen appears.
From here you can:
- View AIE status at-a-glance in the System Health area at the top of the page shows. System Health shows items requiring attention. Click System Health above the graph to access the System Status screen.
- Specify the desired time period for which events appear using the Filter on Timeline area.
- Clear all events with a single comment by clicking Acknowledge All Events.
- Print the filtered list of events by clicking Print.
- Sort/resort the table by the desired data by clicking the desired column heading.
- Hide the navigation pane with the show/hide button to see more or less of the events table as desired.
System Event Information
The unique ID number of the event. Note that ID numbers can cover multiple projects and nodes, so they do not always appear sequential when viewed within a single project.
Check the box to acknowledge the event and to log a comment for it.
This column indicates the *log4j* level of the event.
Indicates the node on which the event occurred.
A brief event description.
The date/time the event occurred.
Indicates where the message originated within AIE.
Large Event Volume
When the number of events go into the tens of thousands, the events are cached in memory within the aie-node to make recall faster. Because the cache is stored in ram, the amount of events the system is capable of retaining is proportional to the maxmem of the aie-node. In other words, if you start running into java heap issues when trying to view system events, you can resolve it by increasing the maxmem of your node
Note that the cache is limited to 200k events, and quick event recall is not guaranteed beyond that point
Since events are stored for only 7 days, having enough events to run afoul of the 200k limit should be considered an unexpected occurrence of great concern, and it's strongly advised that you examine the activities that are responsible for generating the events for any unusual behavior.