Events API

Events API allows you to retrieve events that occurred on your cluster.

The information is only displayed for the queried node.

If you need information on other nodes, it is necessary to query each node separately.

Retrieve all events that happened today

Retrieve events that happened today

CODE

GET /api/monitor/get/events

Since a particular DateTime

Retrieve all events since a specified Unix timestamp. This is particularly useful to get an incremental listing, that displays modifications you made since your last listing.

Each listing will give you a JSON response with a last value:

CODE

GET /api/monitor/get/lastevents/:time

You can use the last value to investigate events in the future.

During a particular day

Retrieve all events at the specified date. The date parameter is in the yyyy-mm-dd format.

CODE

GET /api/monitor/get/events/:date

JSON Format

All events have the same JSON structure for the following endpoints.

JS

{
    "reqs": [
        {
            "type": "CLUSTER_START",
            "date": 1574757198519,
            "duration": 36,
            "comment": ""
        },
        {
            "type": "TABLE_LOADED",
            "date": 1574757199483,
            "duration": 333,
            "comment": "etalon.nyc_yellow_full"
        }
    ],
	"runnings":[
	],
    "last":1627450023823
}

Name	Type	Description
(reqs \| runnings).type	String	Name of the event
(reqs \| runnings).date	Timestamp	Unix timestamp for the start of the event
(reqs \| runnings).duration	Integer	Time it took for the event to execute, if available
(reqs \| runnings).comment	String	A comment associated with the event, if available
last	Long	The timestamp of the current listing

The events list is the following

Name	Comment in the event	Description
CLUSTER_START	None	Cluster starts, tables aren't loaded yet
CLUSTER_STOP	None	Cluster stops, thereby receiving a stop command
NODE_DOWN	Number of the node	A specific node is unreachable
CLUSTER_INIT	None	Cluster is reinitializing (after receiving a command from Monitor or the --init command)
MEMORY_LOW	None	Cluster runs out of memory and tries to unload indexes
TABLE_LOADED	Name of the table	A specific table is loaded and available
HYPER_INDEX_LOADED	Name of the index	A HyperIndex is loaded and available
HYPER_INDEX_UNLOADED	Name of the index	A HyperIndex is unloaded because it is lacking memory
SAVE_METADATA	Number of tables that has been upgraded	Cluster is saving metadata
PRELOAD_FINISHED	Number of indexes loaded	Load of frequently used tables is finished (during the start of the cluster)