[HDFS-7270] Add congestion signaling capability to DataNode write protocol

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 2.7.0
• Component/s: datanode
• Hadoop Flags: Reviewed
• Release Note: Introduced a new configuration dfs.pipeline.ecn. When the configuration is turned on, DataNodes will signal in the writing pipelines when they are overloaded. The client can back off based on this congestion signal to avoid overloading the system.
• Assignee: Haohui Mai
• Reporter: Haohui Mai
• Created: 20/Oct/14 21:52
• Updated: 23/Sep/15 18:16
• Resolved: 05/Feb/15 19:15

Description

When a client writes to HDFS faster than the disk bandwidth of the DNs, it saturates the disk bandwidth and put the DNs unresponsive. The client only backs off by aborting / recovering the pipeline, which leads to failed writes and unnecessary pipeline recovery.

This jira proposes to add explicit congestion control mechanisms in the writing pipeline.

Issue Links

• is related to: HDFS-7748 Separate ECN flags from the Status in the DataTransferPipelineAck
• relates to: HDFS-8009 Signal congestion on the DataNode
• relates to: HDFS-8008 Support client-side back off when the datanodes are congested
• supercedes: HDFS-7265 Use a throttler for replica write in datanode