We are using the Advanced Search adapter for Solr by @Daniel_KM and using a Python script to load several hundred items into Omeka via the API. Each time we update an item, a little background job is created to update the Solr index.
I hadn’t got the Solr setup correct, and perhaps there were other problems with our server configuration. Anyhow, we ended up with PHP running out of memory and various things having to be restarted. We are now left in the Jobs area of Omeka with several hundred indexing jobs which are supposedly “in progress” but which do not have any corresponding process because those all got killed.
The configuration is now all correct, and the little background jobs complete very fast. But we are left with the old “in progress” jobs. If we choose to reindex the whole system via the Search Manager admin interface, it warns us that there are loads of indexing jobs that are still in progress. (The can force it to reindex, but the jobs records remain.)
I would therefore like to find a way of deleting these aborted background jobs from the system or making them as failed, so that this warning does not come up repeatedly. Having searched around, I realise this is not confined to SearchSolr background jobs, as people can easily have other “in progress” jobs that no longer exist.
The jobs do seem to know that the Solr core is not available, but remain in the In Progress state:
