Hello, I have found MATE to be an incredibly useful tool (thanks for all the work on this!). I was wondering if there are any plans to add support for arithmetic operators? For example, it would be useful to be able to calculate such things as the number of retransmits or the total cumulative latency in a TCP connection. If there are not currently plans for this, is there a specific reason? Has it just been a case of lack of resources, or has it been considered but decided it was too difficult for some reason? Thanks, Ryan asked 11 May '16, 07:07 ryber edited 11 May '16, 07:25 Jaap ♦ |
One Answer:
Well, as with most things in Open Source, there isn't really a plan. People contribute what they need or find useful. Apparently no one has thus far had sufficient need for such a functionality so no one has worked on it. Typically I'd consider using the IO Graph for the things you're talking about; I don't quite get how you'd intend to use them in MATE. Anyway if you want to write up a detailed description of what you're looking for (possibly including example MATE configurations) you can always open an enhancement request in case someone picks the idea up and runs with it. (If you do, please report the number back here.) Of course, even better would be code to implement it. :-) answered 11 May '16, 13:13 JeffMorriss ♦ |
Thanks for your response, Jeff.
What I would like to do is be able to easily identify (in a large pcap) TCP streams with a large number of some indicator of dysfunction. For example, any 4-tuples with a large number of TCP resets. Or streams with the most retransmits, etc. And possibly to be able to do some tracking of ack and seq numbers. I have found it to be very powerful in that it can determine whether something has occurred for a given stream, tuple, etc. but it would be even better if it could track the number of occurrences.
At the moment, I am finding that I can do this with Lua. But if I have some time I will try looking through the MATE source code and seeing if it is something I might be able to implement (if so I will certainly let you know :) )
Thanks, Ryan