[SAMZA-153] Grid script should cache downloads

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Assignee: Martin Kleppmann
• Reporter: Martin Kleppmann
• Created: 21/Feb/14 18:02
• Updated: 03/Mar/14 23:37
• Resolved: 21/Feb/14 19:19

Description

At the moment, every time you run bin/grid bootstrap in hello-samza, the dependencies (Samza, Hadoop/YARN, Kafka and Zookeeper) are re-downloaded, even if they have already been downloaded previously. As these downloads are quite large – 100 MB for YARN alone – this process could be much faster if downloads were cached.