Bumping to CloudSuite 4

[xusine]

In order to tune the workloads, we rely on small tweaks.

We should make running and argument passing consistent by :

• [x] Data Caching pass docker_servers.txt for client as an argument instead of modifying it through docker #399
• [x] Data Caching separate dataset and warmup to own container #399
• [x] Data Caching pure load #399
• [x] Data Caching pass argument total number of request to execute instead of infinite #399
• [x] Data Serving separate warmup and load to different commands #398
• [x] Data Serving add command to compact #398
• [x] Media Streaming the log extraction argument
• [x] Media Streaming should print the connection rate.
• [x] Media Streaming friendly log #386
• [x] Web Search separate dataset and uncompression
• [x] Web Search Simlink instead of copying #389
• [x] Web Search the tuning parameters should be given to the docker command
• [x] Web Search, and others Add argument to print or not to print log
• [x] Web Serving: delete the tar file after decompression #421
• [x] Web Serving: don't copy, use symlink or similar instead #421
• [x] Web Serving: merge changes suggested by Mike
• [x] In-memory and graph analytics separate master and worker in documents to match data analytics #416
• [x] Fix the memory requirements for analytic workloads 16GB and if possible 32GB
• [x] #420
• [x] Django workload (deprecated) #397
• [x] For data sets, offer option to either mount from docker image or to mount from folder to control destination directory

[xusine]

Formatting:

• [ ] Putting the entry point of all containers to a fixed position. Now we have entry points like /etc/docker-entrypoint or scripts/docker-entrypoint.