Apache
/
hadoop
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

YARN-1615. Fix typos in delay scheduler's description. Contributed by Akira Ajisaka. 

(cherry picked from commit b8a14efdf5)
This commit is contained in:
Tsuyoshi Ozawa 2015-02-19 10:24:07 +09:00
parent b8fe1a747e
commit fef5efa0cb
2 changed files with 6 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
hadoop-yarn-project/CHANGES.txt
View File

@ -581,6 +581,9 @@ Release 2.7.0 - UNRELEASED
YARN-3132. RMNodeLabelsManager should remove node from node-to-label mapping
when node becomes deactivated. (Wangda Tan via jianhe)
YARN-1615. Fix typos in description about delay scheduling. (Akira Ajisaka via
ozawa)
Release 2.6.0 - 2014-11-18
INCOMPATIBLE CHANGES

6
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FSAppAttempt.java
View File

@ -80,10 +80,10 @@ public class FSAppAttempt extends SchedulerApplicationAttempt
/**
* Delay scheduling: We often want to prioritize scheduling of node-local
* containers over rack-local or off-switch containers. To acheive this
* we first only allow node-local assigments for a given prioirty level,
* containers over rack-local or off-switch containers. To achieve this
* we first only allow node-local assignments for a given priority level,
* then relax the locality threshold once we've had a long enough period
* without succesfully scheduling. We measure both the number of "missed"
* without successfully scheduling. We measure both the number of "missed"
* scheduling opportunities since the last container was scheduled
* at the current allowed level and the time since the last container
* was scheduled. Currently we use only the former.