Part 1: Build M/M/1 model #170

taivop · 2016-12-01T10:38:37Z

Complete by Thu 08 Dec 2016!

TODO:

entire system: from client point of view as one system
if model’s predicted response times are completely wrong, then it is OK? because model is not appropriate for our case?
- is it OK to say model is simply not appropriate?
  - A: possibly. but maybe another problem
say where the data came from (new trace from MS2)
[slack] take peak throughput of the trace for service rate and mean for arrival rate. That's the approach Alonso mentioned in the lecture.

taivop added analysis report labels Dec 1, 2016

taivop added this to the Milestone 3 milestone Dec 1, 2016

taivop added the priority label Dec 1, 2016

taivop mentioned this issue Dec 1, 2016

Part 2: Build M/M/m model #171

Closed

4 tasks

taivop added a commit that referenced this issue Dec 5, 2016

start on #170 and ms3_common.r script

810dc09

taivop added a commit that referenced this issue Dec 5, 2016

work on #170

093a293

taivop added a commit that referenced this issue Dec 6, 2016

working on #170, also #175

5ac211f

taivop added a commit that referenced this issue Dec 6, 2016

#170 some graphs into report

dae2889

taivop added a commit that referenced this issue Dec 6, 2016

#170 everything before comparison

e082bc9

taivop added a commit that referenced this issue Dec 6, 2016

autogenerate table #170

8385ddf

taivop closed this as completed Dec 6, 2016

Provide feedback