[HADOOP-10950] rework heap management vars

• Type: New Feature
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 3.0.0-alpha1
• Component/s: scripts
• Labels: scripts
• Hadoop Flags: Incompatible change

Release Note

<!-- markdown -->

* HADOOP\_HEAPSIZE variable has been deprecated (It will still be honored if set, but expect it to go away in the future). In its place, HADOOP\_HEAPSIZE\_MAX and HADOOP\_HEAPSIZE\_MIN have been introduced to set Xmx and Xms, respectively.

* The internal variable JAVA\_HEAP\_MAX has been removed.

* Default heap sizes have been removed. This will allow for the JVM to use auto-tuning based upon the memory size of the host. To re-enable the old default, configure HADOOP\_HEAPSIZE_MAX="1g" in hadoop-env.sh.

* All global and daemon-specific heap size variables now support units. If the variable is only a number, the size is assumed to be in megabytes.

• Assignee: John Smith
• Reporter: Allen Wittenauer
• Created: 08/Aug/14 19:07
• Updated: 12/May/16 18:22
• Resolved: 10/Dec/14 21:37

Description

Post-HADOOP-9902, we need to rework how heap is configured for small footprint machines, deprecate some options, introduce new ones for greater flexibility.

Issue Links

• duplicates: YARN-115 yarn commands shouldn't add "m" to the heapsize
• is related to: HADOOP-11010 Post-9902 "Umbrella" JIRA
• is related to: HADOOP-11346 Rewrite sls/rumen to use new shell framework
• is related to: YARN-2438 yarn-env.sh cleanup
• relates to: HADOOP-11022 User replaced functions get lost 2-3 levels deep (e.g., sbin)