In 0.7.3, the way of installing spark on mesos was to unpack it into the
same directory across the cluster (I assume this includes the driver
program).  We automated this process in our Ansible templates and all is
right with the world.

In the current 0.8.0 release, the process has been changed implying that
you need to put the Spark tarball in Hadoop and set additional properties
'spark.executor.uri' for the install to work.

The install pattern from 0.7.3 seems to still work and I plan to continue
with this - will it be supported when 0.9.0 stable comes out?  The current
SNAPSHOT still supports the old behavior so just a general question.

Thanks,

Gary

Reply via email to