Other database connection issues
Issue: Data cube job or database query never ends or ends in error
- A job that builds or updates your data cube warehouse or in-memory storage 'hangs' or does not complete. An error message may be reported in the job run details or logs when the job completes or is canceled.
- For certain queries or for affected jobs, an error message containing text such as the following (possibly contained within other messages): A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond
Some connections to your database may have been terminated by network infrastructure such as a firewall after a certain period of time or due to other reasons. Some data providers have a keep alive option that offers a potential solution to this problem:
- For example when editing a data connector using the Oracle Database data provider in Dundas BI version 10 or higher, expand the Advanced section and enable the Keep Alive option. You can read more about the Keep Alive and related Keep Alive Interval and Keep Alive Time settings on Oracle's website.
- Similarly, you can edit a data connector using the PostgreSQL data provider in Dundas BI version 10 or higher, and expand the Performance section to find the Keepalive and TCP Keepalive options. Each of these options is used in different scenarios. For details, see the Npgsql page on the Keepalive options.
For data cube storage jobs in Dundas BI 10.0.0.1002 and higher, there is a configuration setting named Database Bulk Copy Timeout, found in the General category with advanced settings shown. You may prefer to set this to a shorter amount of time when a job takes unnecessarily long to prevent it from running continuously in case one of its copy operations is stalled.