Saved Search: make links using a timer (and/or publish triggers) rather than curl and cron #2543
Labels
Comments
This was referenced Nov 19, 2015
pdurbin
added a commit
that referenced
this issue
Sep 15, 2016
In a refactoring at c10cbd9 we lost the meaning of 0.0.0.0 which was "Setting address to undefined (0.0.0.0)" so this commit makes the code clearer. A note has been added that Saved Search may not play nicely with IP Groups but this should probably be addressed along with the long list of other Saved Search issues mentioned in the API Guide: #1364, #1813, #1840, #1890, #1939, #2167, #2186, #2053 and #2543
|
Meh, I like cron. And EJB timers seem to be problematic: http://guides.dataverse.org/en/4.7/admin/troubleshooting.html#deployment-fails-ejb-timer-service-not-available . Closing. |
|
@pdurbin, is the cron job/ need for the cron job documented in the guides? |
|
@kcondon yes, the need to use cron or similar is documented at http://guides.dataverse.org/en/4.9.2/api/native-api.html#saved-search |
|
@pdurbin Seems kind of buried and easy to overlook. Anyway, no skin off my nose. I'm not sure even we run a cron job. |
|
Maybe instead of "documented" I should have said "mentioned in passing." 😄 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
In #2531 we are working on documenting how to set up curl and cron to kick over periodic linking of dataverses and datasets based on saved searches but longer term we'd like this to be part of the app itself.
We should think about how often the equivalent of the "makelinks" API endpoint(s) should be called and if we'll let non-superusers choose how often.
As @sbarbosadataverse pointed out, it would also be nice if the linking were triggered but "publish" actions of any dataset in a dataverse.
The text was updated successfully, but these errors were encountered: