Create New SQL Data Source
You can use a JDBC driver to pull data into Metric Insights using a SQL fetch command. This includes traditional databases as well as NoSQL sources with JDBC access; for example, Hadoop Hive.
This article describes the general process for creating a SQL data source. Information on creating a data sources based on a plugin can be found Create New SQL Data Source.
For a description of Metric Insights overall approach to Data Sources, click Understanding Data Sources.
1. Access Admin > Data Sources
At the bottom of the page click [+New Data Source]
New in 6.2.2:
Filters added for Type and Created by
Columns added for Certified? and Created by
2. Choose type
- Select SQL as the type of a new Data Source
- Next step
2.1. Complete all settings
- The Port number will be set by default, based on your choice of JDBC Driver. Change it if necessary.
- The JDBC string will be created automatically based on your other inputs. In some cases, however, it will not be possible to infer the correct string without additional inputs. If the Connection Test fails, contact Metric Insights for assistance..
- Optionally, specify the maximum number of concurrent Threads per Trigger execution to be used in background processing when the system updates Metrics and Reports for this Data Source. If you do not specify any value for this setting, batch data collection processing will be single-threaded.
- If you select 'yes' to specify that this Data Source is remote; for example, behind a firewall, you will be required to select a Connector or create a new one. See Configuring a Remote Data Processor (v6+) for more details.
- Once a new Data Source is saved, click Test Connection to ensure that settings are correct. NOTE: Some data sources cannot be tested this way.
2.2. Review and close the confirmation
If the connectivity is established, the confirmation message appears; click OK to continue.
NOTE: It is not possible to directly test Hive connectivity from the connection editor since there is no standard query that can be run against a HiveQL instance. Contact Metric Insights for assistance in validating a HiveQL connection.