You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
At the moment, there is a gap of a few seconds between clearing a index and the new content being uploaded and indexed.
That probably isn't a problem, but if something in the new data is rejected by Solr (e.g. bodleian/medieval-mss#118) then the web site will lack that index (most likely the manuscript one, so the manuscript view pages will be blank too) until it is fixed. Again, if the practice of testing first on QA before loading to production is followed (as Matthew did in that example) then it is fine, but it would be good to make it foolproof.
The proposed solution is to copy the setup on Digital.Bodleian and keep two index cores on solr01-prd, and have a third step in the indexing script that only swaps them if the first two steps don't return errors.
The text was updated successfully, but these errors were encountered:
At the moment, there is a gap of a few seconds between clearing a index and the new content being uploaded and indexed.
That probably isn't a problem, but if something in the new data is rejected by Solr (e.g. bodleian/medieval-mss#118) then the web site will lack that index (most likely the manuscript one, so the manuscript view pages will be blank too) until it is fixed. Again, if the practice of testing first on QA before loading to production is followed (as Matthew did in that example) then it is fine, but it would be good to make it foolproof.
The proposed solution is to copy the setup on Digital.Bodleian and keep two index cores on solr01-prd, and have a third step in the indexing script that only swaps them if the first two steps don't return errors.
The text was updated successfully, but these errors were encountered: