[HDFS-6016] Update datanode replacement policy to make writes more robust

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Won't Fix
• Component/s: datanode, ha, hdfs-client, namenode
• Target Version/s: HDFS-5535 (Rolling upgrades)
• Assignee: Kihwal Lee
• Reporter: Kihwal Lee
• Created: 25/Feb/14 17:48
• Updated: 13/Aug/14 15:45
• Resolved: 13/Aug/14 15:45

Description

As discussed in HDFS-5924, writers that are down to only one node due to node failures can suffer if a DN does not restart in time. We do not worry about writes that began with single replica.