User login

Weekly Report

23

Aug

2016

I'm working on the first stages of placing rules based on the dependency graph. As a starting point I'm looking at narrowing down possible placements for individual rules as quickly as possible. To begin with I'm taking all possible placement of every rule, then for each dependency keeping a list of those locations that satisfy the original dependency.

Any single rule may have a number of dependencies, with each dependency being satisfied by a subset of possible placements. The common placements across dependencies are then chosen as the best starting points. Currently only single dependencies are being considered, I will extend this to consider the entire tree. I'm currently not considering possible types of traffic that will hit a given rule, such may exclude portions of traffic from hitting a rule which could be that portion that is problematic to place.

I also tidied the code base a bit and added everything back into git. After splitting a base TTP library out from the matching portion of the code and from the executable tools a couple of weeks ago, I had not got around to adding these back to git. I also split up a couple of the tools, the flow collection tool is now standalone and will collect every possible metric it can query from OpenFlow, including groups which I originally forgot to collect and other info a switch will present.