aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorNicholas Chammas <nicholas.chammas@gmail.com>2014-11-05 20:45:35 -0800
committerShivaram Venkataraman <shivaram@cs.berkeley.edu>2014-11-05 20:45:35 -0800
commitdb45f5ad0368760dbeaa618a04f66ae9b2bed656 (patch)
tree9e6916d5a46ea9d52d36e6862d37fc55d639df1b
parent3d2b5bc5bb979d8b0b71e06bc0f4548376fdbb98 (diff)
downloadspark-db45f5ad0368760dbeaa618a04f66ae9b2bed656.tar.gz
spark-db45f5ad0368760dbeaa618a04f66ae9b2bed656.tar.bz2
spark-db45f5ad0368760dbeaa618a04f66ae9b2bed656.zip
[SPARK-4137] [EC2] Don't change working dir on user
This issue was uncovered after [this discussion](https://issues.apache.org/jira/browse/SPARK-3398?focusedCommentId=14187471&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14187471). Don't change the working directory on the user. This breaks relative paths the user may pass in, e.g., for the SSH identity file. ``` ./ec2/spark-ec2 -i ../my.pem ``` This patch will preserve the user's current working directory and allow calls like the one above to work. Author: Nicholas Chammas <nicholas.chammas@gmail.com> Closes #2988 from nchammas/spark-ec2-cwd and squashes the following commits: f3850b5 [Nicholas Chammas] pep8 fix fbc20c7 [Nicholas Chammas] revert to old commenting style 752f958 [Nicholas Chammas] specify deploy.generic path absolutely bcdf6a5 [Nicholas Chammas] fix typo 77871a2 [Nicholas Chammas] add clarifying comment ce071fc [Nicholas Chammas] don't change working dir
-rwxr-xr-xec2/spark-ec28
-rwxr-xr-xec2/spark_ec2.py12
2 files changed, 17 insertions, 3 deletions
diff --git a/ec2/spark-ec2 b/ec2/spark-ec2
index 31f9771223..4aa908242e 100755
--- a/ec2/spark-ec2
+++ b/ec2/spark-ec2
@@ -18,5 +18,9 @@
# limitations under the License.
#
-cd "`dirname $0`"
-PYTHONPATH="./third_party/boto-2.4.1.zip/boto-2.4.1:$PYTHONPATH" python ./spark_ec2.py "$@"
+# Preserve the user's CWD so that relative paths are passed correctly to
+#+ the underlying Python script.
+SPARK_EC2_DIR="$(dirname $0)"
+
+PYTHONPATH="${SPARK_EC2_DIR}/third_party/boto-2.4.1.zip/boto-2.4.1:$PYTHONPATH" \
+ python "${SPARK_EC2_DIR}/spark_ec2.py" "$@"
diff --git a/ec2/spark_ec2.py b/ec2/spark_ec2.py
index 50f88f7356..a5396c2375 100755
--- a/ec2/spark_ec2.py
+++ b/ec2/spark_ec2.py
@@ -40,6 +40,7 @@ from boto.ec2.blockdevicemapping import BlockDeviceMapping, BlockDeviceType, EBS
from boto import ec2
DEFAULT_SPARK_VERSION = "1.1.0"
+SPARK_EC2_DIR = os.path.dirname(os.path.realpath(__file__))
MESOS_SPARK_EC2_BRANCH = "v4"
# A URL prefix from which to fetch AMI information
@@ -593,7 +594,14 @@ def setup_cluster(conn, master_nodes, slave_nodes, opts, deploy_ssh_key):
)
print "Deploying files to master..."
- deploy_files(conn, "deploy.generic", opts, master_nodes, slave_nodes, modules)
+ deploy_files(
+ conn=conn,
+ root_dir=SPARK_EC2_DIR + "/" + "deploy.generic",
+ opts=opts,
+ master_nodes=master_nodes,
+ slave_nodes=slave_nodes,
+ modules=modules
+ )
print "Running setup on master..."
setup_spark_cluster(master, opts)
@@ -730,6 +738,8 @@ def get_num_disks(instance_type):
# cluster (e.g. lists of masters and slaves). Files are only deployed to
# the first master instance in the cluster, and we expect the setup
# script to be run on that instance to copy them to other nodes.
+#
+# root_dir should be an absolute path to the directory with the files we want to deploy.
def deploy_files(conn, root_dir, opts, master_nodes, slave_nodes, modules):
active_master = master_nodes[0].public_dns_name