[ https://issues.apache.org/jira/browse/HDFS-1973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13092073#comment-13092073 ]
Aaron T. Myers commented on HDFS-1973: -------------------------------------- @Eli, sure I'll file a separate JIRA. It'd certainly be worth enumerating all of the places where HTTP fail-over is an issue. The example you provided is an interesting one. It seems you're assuming that an HA setup would have three nodes - active, standby, and 2NN, with the 2NN failing over to do checkpointing against the standby after a failure of the active. The design document in HDFS-1623 doesn't really address checkpointing. I've heard from Suresh and Todd informally that the intention is probably to make the standby node also capable of performing checkpointing. I'll file a separate JIRA to address this as well. > HA: HDFS clients must handle namenode failover and switch over to the new > active namenode. > ------------------------------------------------------------------------------------------ > > Key: HDFS-1973 > URL: https://issues.apache.org/jira/browse/HDFS-1973 > Project: Hadoop HDFS > Issue Type: Sub-task > Reporter: Suresh Srinivas > Assignee: Aaron T. Myers > > During failover, a client must detect the current active namenode failure and > switch over to the new active namenode. The switch over might make use of IP > failover or some thing more elaborate such as zookeeper to discover the new > active. -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira