[MAPREDUCE-240] Improve the shuffle phase by using the "connection: keep-alive" and doing batch transfers of files

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Duplicate
• Assignee: Jothi Padmanabhan
• Reporter: Devaraj Das
• Created: 08/May/07 06:17
• Updated: 11/Sep/09 04:25
• Resolved: 11/Sep/09 04:25

Description

We should do transfers of map outputs at the granularity of total-bytes-transferred rather than the current way of transferring a single file and then closing the connection to the server. A single TaskTracker might have a couple of map output files for a given reduce, and we should transfer multiple of them (upto a certain total size) in a single connection to the TaskTracker. Using HTTP-1.1's keep-alive connection would help since it would keep the connection open for more than one file transfer. We should limit the transfers to a certain size so that we don't hold up a jetty thread indefinitely (and cause timeouts for other clients).
Overall, this should give us improved performance.

Issue Links

• is cloned by: HADOOP-3656 fetcher should re-use connection when it needs to fetch multiple segments from the same task tracker