[ https://issues.apache.org/jira/browse/NIFI-1452?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16017931#comment-16017931 ]
Michael Moser commented on NIFI-1452: ------------------------------------- This also affects GetHTTP, which yields when HTTP server responds with 304 NOT_MODIFIED. So a GetHTTP with Run Schedule of 10 minutes and Yield Duration of 1 second, actually hits the HTTP server every 1 second when the server replies with 304. An example server that exhibits this behavior is http://archive.apache.org/icons/blank.gif > Yield Duration can short circuit long Timer Driven Run Schedule > --------------------------------------------------------------- > > Key: NIFI-1452 > URL: https://issues.apache.org/jira/browse/NIFI-1452 > Project: Apache NiFi > Issue Type: Bug > Components: Core Framework > Affects Versions: 0.4.1 > Reporter: Michael Moser > Assignee: Michael Moser > Priority: Trivial > > This may be a rare use case, but I configured a GetFile processor to be Timer > Driven with a Run Schedule of 30 secs. Its Yield Duration was default 1 sec. > I expected GetFile onTrigger() to be called every 30 secs, but it was being > called every 1 sec most of the time. > GetFile will call context.yield() when it polls a directory and gets an empty > list in return. It appears that a yield will ignore the Run Schedule. Many > standard processors call context.yield() when they have no work to do. > I changed my scheduling strategy to CRON Driven with its run schedule every > 30 seconds, and the onTrigger() was called every 30 seconds, even after a > yield. So CRON Driven scheduling is working as expected after a yield. -- This message was sent by Atlassian JIRA (v6.3.15#6346)