aboutsummaryrefslogtreecommitdiff
path: root/ec2
diff options
context:
space:
mode:
authorShivaram Venkataraman <shivaram@cs.berkeley.edu>2015-05-26 15:01:27 -0700
committerAndrew Or <andrew@databricks.com>2015-05-26 15:01:27 -0700
commit2e9a5f229e1a2ccffa74fa59fa6a55b2704d9c1a (patch)
tree8aca19bcc06903a77d999dad4b4fdb1a5237439a /ec2
parent61664732b25b35f94be35a42cde651cbfd0e02b7 (diff)
downloadspark-2e9a5f229e1a2ccffa74fa59fa6a55b2704d9c1a.tar.gz
spark-2e9a5f229e1a2ccffa74fa59fa6a55b2704d9c1a.tar.bz2
spark-2e9a5f229e1a2ccffa74fa59fa6a55b2704d9c1a.zip
[SPARK-3674] YARN support in Spark EC2
This corresponds to https://github.com/mesos/spark-ec2/pull/116 in the spark-ec2 repo. The only changes required on the spark_ec2.py script is to open the RM port. cc andrewor14 Author: Shivaram Venkataraman <shivaram@cs.berkeley.edu> Closes #6376 from shivaram/spark-ec2-yarn and squashes the following commits: 961504a [Shivaram Venkataraman] Merge branch 'master' of https://github.com/apache/spark into spark-ec2-yarn 152c94c [Shivaram Venkataraman] Open 8088 for YARN in EC2
Diffstat (limited to 'ec2')
-rwxr-xr-xec2/spark_ec2.py2
1 files changed, 2 insertions, 0 deletions
diff --git a/ec2/spark_ec2.py b/ec2/spark_ec2.py
index 724811eaa1..ee0904c9e5 100755
--- a/ec2/spark_ec2.py
+++ b/ec2/spark_ec2.py
@@ -491,6 +491,8 @@ def launch_cluster(conn, opts, cluster_name):
master_group.authorize('udp', 2049, 2049, authorized_address)
master_group.authorize('tcp', 4242, 4242, authorized_address)
master_group.authorize('udp', 4242, 4242, authorized_address)
+ # RM in YARN mode uses 8088
+ master_group.authorize('tcp', 8088, 8088, authorized_address)
if opts.ganglia:
master_group.authorize('tcp', 5080, 5080, authorized_address)
if slave_group.rules == []: # Group was just now created