Database Reference
In-Depth Information
Further Reading
This chapter has given a short overview of YARN. For more detail, see
Apache Hadoop
YARN
by Arun C. Murthy et al. (Addison-Wesley, 2014).
[
35
]
It's also possible for the client to start the application master, possibly outside the cluster, or in the same
JVM as the client. This is called an
unmanaged application master
.
[
36
]
For more on this topic see
Scaling Out
and
Network Topology and Hadoop
.
[
37
]
The low-latency application master code lives in the
Llama project
.
[
39
]
As of Hadoop 2.5.1, the YARN timeline server does not yet store MapReduce job history, so a MapRe-
duce job history server daemon is still needed (see
Cluster Setup and Installation
)
.
[
40
]
Arun C. Murthy,
“The Next Generation of Apache Hadoop MapReduce,”
February 14, 2011.
a value larger than 1 (the default), then a single job is allowed to use more than its queue's capacity.
[
42
]
However, the Capacity Scheduler can perform work-preserving preemption, where the resource manager
asks applications to return containers to balance capacity.
[
43
]
DRF was introduced in Ghodsi et al.'s
“Dominant Resource Fairness: Fair Allocation of Multiple Re-
source Types,”
March 2011.