Change Details

As an engineering manager, I do not want to unnecessarily access origin servers for Related Articles, but instead want edge cached responses, so that the origin server capacity can be used for other purposes. As a reader, I want Related Articles to load quickly, so that I can more easily find other neat stuff at a glance. This is the bridge solution that is a precursor to {T125983}. The api.php `smaxage` parameter should be set on the Related Articles `morelike` requests with a value of 24 hours. ##Acceptance criteria [ ] The solution should not impinge on the freshness of human generated suggestions if they are present. It is assumed that updated human generated suggestions are reflected in updated parser output already (i.e., human generated suggestions are treated preferentially), but this is worth verification. [ ] It works on the beta cluster. [ ] It is slated for the production cluster (presently Related Articles is only available in the beta user mode of the production cluster).

As an engineering manager, I do not want to unnecessarily access origin servers for Related Articles, but instead want edge cached responses, so that the origin server capacity can be used for other purposes. As a reader, I want Related Articles to load quickly, so that I can more easily find other neat stuff at a glance. This is the bridge solution that is a precursor to {T125983}. The api.php `smaxage` parameter should be set on the Related Articles `morelike` requests with a value of 24 hours.

As an engineering manager, I do not want to unnecessarily access origin servers for Related Articles, but instead want edge cached responses, so that the origin server capacity can be used for other purposes. As a reader, I want Related Articles to load quickly, so that I can more easily find other neat stuff at a glance. This is the bridge solution that is a precursor to {T125983}. The api.php `smaxage` parameter should be set on the Related Articles `morelike` requests with a value of 24 hours. ##Acceptance criteria [ ] The solution should not impinge on the freshness of human generated suggestions if they are present. It is assumed that updated human generated suggestions are reflected in updated parser output already (i.e., human generated suggestions are treated preferentially), but this is worth verification. [ ] It works on the beta cluster. [ ] It is slated for the production cluster (presently Related Articles is only available in the beta user mode of the production cluster).