[ https://issues.apache.org/jira/browse/HBASE-20847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16537045#comment-16537045 ]
stack edited comment on HBASE-20847 at 7/9/18 3:15 PM: ------------------------------------------------------- bq. The locks which have already been held by parent procedures should also be considered as held by sub procedures. This is OK as we can make sure that the parent procedure will not release the lock before the sub procedures, as it can only be executed again after all the sub procedures have finished. Are we talking about the post-master crash scenario? Is the suggestion that sub-procedures just assume and put back in place the locks that were taken out by parent procedures when they run post-crash? And this will be ok as long as the parent waits on all sub-procedures to complete before it lets-go of shared or exclusive lock? (Which is how it currently works?). If a sub-procedure is scheduled post-crash, why not wait and let the parent procedure run again so it can create any locks? Is it because parent procedures are not written to do this? Because parents cede control to sub-procedures and depend on their completion before they can be scheduled again? (I can help by writing this up... Thanks) was (Author: stack): bq. The locks which have already been held by parent procedures should also be considered as held by sub procedures. This is OK as we can make sure that the parent procedure will not release the lock before the sub procedures, as it can only be executed again after all the sub procedures have finished. Are we talking about the post-master crash scenario? Is the suggestion that sub-procedures assume locks that were taken out by parent procedures? > The parent procedure of RegionTransitionProcedure may not have the table lock > ----------------------------------------------------------------------------- > > Key: HBASE-20847 > URL: https://issues.apache.org/jira/browse/HBASE-20847 > Project: HBase > Issue Type: Sub-task > Components: proc-v2, Region Assignment > Reporter: Duo Zhang > Assignee: Duo Zhang > Priority: Major > Attachments: HBASE-20847-v1.patch, HBASE-20847.patch > > > For example, SCP can also schedule AssignProcedure and obviously it will not > hold the table lock. -- This message was sent by Atlassian JIRA (v7.6.3#76005)