Nagios

Overview

The glue server returns nagios compatible information about each workflow execution. This means that nagios can be used easily to monitor and notify when workflows have failed.

Monitoring Workflows

Checks will return

0 | if the service is ok
2 | if the service has error or did not run with in the specified time period
3 | if the workflow cannot be found in the glue repository

Nagios can either check for single workflows i.e.

http://<gluesever>:<port>/check/<workflowName>?expectedRun=30

This will check the status of the last run for the workflow specified and return:

2 | if the last run was ERROR OR the last run was completed more than 30 minutes ago.
0 | if the last run was SUCCESS AND the last run was completed less than 30 minutes ago.

More information can be supplied via the info call:

http://<gluesever>:<port>/history/<workflowName>?limit=10

This will return the last 10 runs one line per run in the following format:

Unit Run ID,STATUS(SUCCESS|ERROR),start date time,end  date time

The date time fields are formatted using yyyy-MM-dd HH:mm:ss

Examples

Did workflow mytest run with a status OK during the last 20 minutes

http://localhost:8025/check/mytest?expectedRun=20

Did workflow mytest run with a status OK during the last 24 hours (60 * 24) minutes

http://localhost:8025/check/mytest?expectedRun=1440

History last 5 days' runs of mytest

http://localhost:8025/history/mytest?limit=5

Glue