cs-6381/finalproject

finalproject's Issues

Add Flume and Pulsar experiments

Private or Public IP

It's unclear whether or not we should use the private or public IP addresses on Chameleon Cloud.

In class, @TheENGR stated that it would be better to use the private IP address to avoid contamination from outside sources.

Clarification on Constants

1. type/specs of machine running (Ubuntu 18.04 chameleon cloud VM’s with 8 CPU and 32 GB of RAM — but chameleon cloud does not allow 8 CPUs and 32 gb of ram so.. we should use the 8 CPU and 16 gb of ram instead?)
2. number of topics (1 — MQ’s without ACK - if we make the producer also be a consumer and have it listen for messages with its ID will this not add another topic?)
3. number of brokers (still unclear) - we just have one VM for one instance of broker/system, correct?

Performance MQ Blended Test

The experiment configuration for this makes no sense. We should be varying the pub/sub configuration for each message size. Theoretically you would have the first 5 iterations from test 1 and then multiply them for each iteration/variation in test 2 (5 iterations). This would result in 25 tests. You could lower this number by 5 because the first test has a message size of “small”. I personally don’t think we would need 25 tests… I would recommend running more of the extreme cases like tiny, medium, x large. And maybe even cut out some of the test 1 tests (i.e. just us the first 3 cases involving 1 machine per consumer/producer).

Dependent Variables - clarification on csv files

Data processor:

• does “experiment designers” refer to the people running the tests or the actual experiment designers?
• Is the first .csv “processing latency” the roundtrip time for the message to go from pub to consumer and back? If not, wouldn’t we have. processing latency for both the pub and the sub? Which would we care about (if not both). In other words (just to be suppppperrrr clear) Do we want End to End latency for pub -> sub AND end to end latency for sub -> pub response?
• wouldn’t the second csv file also be two separate files? One for the number of messages sent per second by the pub. another for the number of messages received by the sub?