[ https://issues.apache.org/jira/browse/GIRAPH-297?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13433377#comment-13433377 ]
Jan van der Lugt commented on GIRAPH-297: ----------------------------------------- This is very difficult, since aggregator values are collected and immediately sent out again at the end of each superstep. The master.compute has to fit in between these two step (I actually had to cut up a function called collectAndProcessAggregators). I cannot imagine another point in time when it can be executed. So if master.compute executes between these steps and we want it (at least I wanted it to) logically execute before the vertices in any given step, this seems (or seemed, now) the best way to do it. > Checkpointing on master is done one superstep later > --------------------------------------------------- > > Key: GIRAPH-297 > URL: https://issues.apache.org/jira/browse/GIRAPH-297 > Project: Giraph > Issue Type: Bug > Reporter: Maja Kabiljo > Assignee: Maja Kabiljo > > On workers we store checkpoint X before compute() for superstep X are > executed. On master we do it after those compute() are executed and after > master.compute() for superstep X+1. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira