Activate datasource timeout

 


Problem

Certain conditions can affect the performance of your datasource, potentially impacting how quickly Elements Connect fields and Jira pages load:

  • High volume of data: If the datasource returns too many results, it can slow down response times
  • Network delays: Any lag between the Jira instance and the datasource can cause a delay in data retrieval
  • Server response times: A datasource hosted on an underperforming server might respond more slowly than desired

Solution

To maintain consistent performance and manage potential data retrieval issues effectively, the Read Timeout feature is available and adjustable for various datasource types:

  • Database
  • LDAP
  • JNDI
  • URL (note: cannot be disabled)

To enhance system stability and prevent performance lags, it's advisable to activate the timeout feature in your datasource settings. This introduces two types of timeouts:

  • Connection Timeout: Fixed at 5 seconds, this cannot be modified
  • Read Timeout: Adjustable by an administrator, typically set from 10 to 30 seconds to suit your needs

When a timeout limit is reached, the system will automatically halt the query to prevent long waits and potential system strain.

To set this up:

  1. Navigate to the Elements Connect administration menu
  2. Select the 'Datasources' tab
  3. Choose and edit your datasource
  4. Turn on the 'Timeout' option
  5. Specify your desired 'Read Timeout Duration'



This solution prevents problems caused by third-party factors from affecting Jira's stability, but it doesn't solve the source of the problem. A thorough investigation is recommended to find and resolve the real cause:

  • If the problem is caused by the datasource trying to return too many results, then you should optimize the Connect field query and add filters to limit the number of results.
  • If the problem is caused by network issues, then this is outside our scope and you should contact your network administrator to investigate.
  • If the problem is caused by your datasource being too slow, then this is outside our scope and you should contact your datasource administrator to investigate.