Hi Greg, >> One of the things we talked about at the Conference in Oxford was >> getting better documentation for the workflow of the capture agent. >> Basically, which REST endpoints are used, when to call the different >> REST endpoints...how often you would call them. The basics needed to >> implement the CA interface. The same stuff you can get by reverse >> engineering and wiresharking the CA, but you know in a document. ;) > > Ok, first pass is done and posted at > > http://opencast.jira.com/wiki/display/MH/Capture+Agent+Communication+Protocols > > Please let me know if anyone thinks there's anything missing or unclear. > This is off the top of my head (since I wrote pretty much all of it) so > it's quite possible I glossed over something.
Thank you for writing this piece of documentation. It make things much clearer. I still have one question though: What is the proper state of an agent _after_ it has been shut down? Is it UNKNOWN? Or should I better remove the capture agent registration (DELETE $HOST/capture-admin/agents/$AGENT_NAME). Without any knowledge of the Matterhorn internals it's not clear what implications the agent states have. Are they only meant to be displayed in the admin UI or are there any parts of MH relying on the state of an agent? -- Basil _______________________________________________ Matterhorn mailing list [email protected] http://lists.opencastproject.org/mailman/listinfo/matterhorn To unsubscribe please email [email protected] _______________________________________________
