Address missing sitemap.xml

[kilahimm]

Investigate best method to ensure sitemap.xml is created on deployment. Cron job does not appear to run when file does not already exit.

[kerchner]

Discoveries so far: On test.scholarspace, the logs show that the cron job ran, but there was no sitemap.xml file. After running the job once manually, and then deleting sitemap.xml, subsequent runs of the cron task are successful, so I haven't been able to reproduce the issue on anything but a new (re)deployment. Note that cron outputs are being discarded because an MTA (email/messaging) server has not been set up; this would typically email the outputs of a cron job. I modified the cron job to redirect to a file (it will take more investigating to ensure the outputs show in the container's docker log). The job succeeded and the outputs were written to the file, but no errors.

Note that running https://github.com/gwu-libraries/scholarspace-hyrax/blob/master/lib/tasks/schoolie.rake#L28 is successful, so it's unlikely that file permission issues are involved.

Removing the app-hyrax docker image and redeploying did not seem to reproduce the problem (may be worth another try). I did not yet try removing the volume itself.

Possible avenues for further investigation:

• Set up a simple docker container with a cron job that writes to a file; confirm there are no inherent issues.
• Modify the docker setup script to touch sitemap.xml and/or to run the sitemap_generate_queue rake task immediately.

More work to be done in order to solve this.