On Sun, Jun 11, 2017 at 8:05 AM, John D. Ament <johndam...@apache.org> wrote: > On Sat, Jun 10, 2017 at 11:51 AM sebb <seb...@gmail.com> wrote: > >> On 10 June 2017 at 15:52, John D. Ament <johndam...@apache.org> wrote: >> > I think there's a bug with how the caching works for podlings.xml, but >> I'm >> > not sure if it's something I introduced or something I'm just noticing. >> > >> > I would expect that the cache is comparing podlings.json to podlings.xml. >> > If podlings.json mod time is newer than podlings.xml, read the values >> from >> > there. Otherwise, refresh from the xml file. >> > >> > Presently, it looks like we're checking the cache folder for the xml file >> > and re-reading that. >> > >> > Thoughts? >> >> I'm not sure what you are referring to here. >> > > https://github.com/apache/whimsy/blob/master/lib/whimsy/asf/podlings.rb#L304 > > I'm not sure if I'm not reading this properly or what (granted I'm a ruby > n00b). It seems like all this is doing is reading podlings.xml from a > different location on the file system instead of leveraging a pre-built > cache of the podling structure. Basically, I don't understand what > performance boost this cache is giving us in the current structure.
That's not reading podlings.xml; that's issuing a request to JIRA to list all of the issues that have ever been raised against the podlingnamesearch project. Those types of requests take time and resources across multiple machines. This information is used by, for example, the board agenda tool. So instead of sending a request each time the board agenda is loaded, it will send a request at most every five minutes. >> > John - Sam Ruby
