Problem
Adam-submit is using spark in the localhost, but if we want yo use another spark cluster, adam-submit is failing.
From what I found out, adam-submit (which is calling spark-submit) don't have ways to give args to the spark-submit.
This is the command i used :
(you need a spark master and a hdfs cluster to run it)
docker run --link=spark-master:spark-master --link=hdfs-namenode:hdfs-namenode gelog/adam \
adam-submit --master spark://spark-master:7077 \
transform \
hdfs://hdfs-namenode:9000/SRR062634.sam \
hdfs://hdfs-namenode:9000/SRR062634.adam
Solution
We cloud tweak the adam-submit script to pass an ENV variable when it call the spark-submit
The env var cloud be something like $SPARK_SUBMIT_ARGS with an empty default value but we cloud gave a value like
SPARK_SUBMIT_ARGS='--master spark://spark-master:7077'
and use the following command:
docker run -e "SPARK_SUBMIT_ARGS=--master spark://spark-master:7077" \
--link=spark-master:spark-master --link=hdfs-namenode:hdfs-namenode gelog/adam \
adam-submit \
transform \
hdfs://hdfs-namenode:9000/SRR062634.sam \
hdfs://hdfs-namenode:9000/SRR062634.adam
Problem
Adam-submit is using spark in the localhost, but if we want yo use another spark cluster, adam-submit is failing.
From what I found out, adam-submit (which is calling spark-submit) don't have ways to give args to the spark-submit.
This is the command i used :
(you need a spark master and a hdfs cluster to run it)
Solution
We cloud tweak the adam-submit script to pass an ENV variable when it call the spark-submit
The env var cloud be something like $SPARK_SUBMIT_ARGS with an empty default value but we cloud gave a value like
and use the following command: