[ https://issues.apache.org/jira/browse/PHOENIX-3045?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15365862#comment-15365862 ]
Ankit Singhal commented on PHOENIX-3045: ---------------------------------------- Workaround could be that the user can create a ASYNC index again or just a index physical table with the same name ,so that the index writes start succeeding to let WAL replay completes , once all data regions are up, then he can drop table after doing flush on the data table. [~lhofhansl],[~enis], let me know if workaround is not that suitable so that I can make this blocker for the 4.8 release. > Data regions in transition forever if RS holding them down during drop index > and before flush on data table > ------------------------------------------------------------------------------------------------------------ > > Key: PHOENIX-3045 > URL: https://issues.apache.org/jira/browse/PHOENIX-3045 > Project: Phoenix > Issue Type: Bug > Reporter: Sergio Peleato > Assignee: Ankit Singhal > Fix For: 4.9.0 > > > Currently we are flushing data table after dropping the index so that region > opening won't be failed to recover dropped index edits. But there is a chance > that region server holding the data regions might abruptly killed before > flushing the data table this leads same failure case that data regions won't > be opened which leads to the regions in transition forever. We need to handle > this case by checking dropped indexes on recovery write failures and skip the > corresponding mutations to write to them. -- This message was sent by Atlassian JIRA (v6.3.4#6332)