[ https://issues.apache.org/jira/browse/CASSANDRA-14231?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jeff Jirsa resolved CASSANDRA-14231. ------------------------------------ Resolution: Invalid > After ddl cluster.getMetadata().checkSchemaAgreement() returns true but > schema is not in agreement > -------------------------------------------------------------------------------------------------- > > Key: CASSANDRA-14231 > URL: https://issues.apache.org/jira/browse/CASSANDRA-14231 > Project: Cassandra > Issue Type: Bug > Components: Core > Reporter: George Shuklin > Priority: Major > Fix For: 3.10 > > > We run a set of ddl statements and after each of them wait until schema come > to agreement. However periodically after Cluster says that schema is in > agreement and we run next DDL statement, it fails as previous DDL did not > work properly. > The most typical scenario when this problem happens almost always: first DDL > drops a table, and next DDL creates it (and cassandra says that this table > already exists). > > This is our code to wait: > private void waitForSchemaAgreement() > { > while (!icluster.getMetadata().checkSchemaAgreement()) { > try { > Thread.sleep(waitTime*1000); > } catch (Exception e){ > // ignore } } } > -- This message was sent by Atlassian JIRA (v7.6.3#76005) --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org