User login

Blogs

07

Jun

2013

Spent this week finishing of the last assignment for 513, so not much work done on my project this week.

07

Jun

2013

This week has been particularly short for me as I came down with a cold on Monday and it plagued me all week. I did manage to merge my code with the latest release of ns-3 with no problems. I also had a go at trimming ns-3 of modules I don't particularly need such as LTE and WiMAX but it seems most of the included modules are tightly integrated with the whole system and it doesn't like it when you fiddle around. I have also been getting some presentation material in place for the inevitable WAND talk I will have to give at some point.

05

Jun

2013

Finished up merging events and properly marking them on the time series
graphs. Need to keep an eye on them now and see how they work in practice
with actual event frequencies and use cases.

Sat down with Shane and had a think about where we are with the
NNTSC/eventing/alerting/etc code and what needs to be done to have it in a
usable state for Lightwire to test. Started planning the path towards that
milestone and creating tickets to do so.

Made lots of small fixes to the new amplet code to remove dead code,
install files in more sensible (and consistent) locations, improve user
control of logging and improve the quality of output from the icmp/dns
tests. After talking with Brad about machine provisioning systems, puppet,
remote management etc decided it was a good idea to move schedule
configuration into a directory rather than using a single file, similar to
how apache, cron and others can work.

04

Jun

2013

After getting an ssh connection issue fixed, the Caida and planetlab scamper runs were kicked off. Planetlab has one node where scamper does not continue, however there are five Caida nodes where this happens. On further analysis I found some driver cores, so I recompiled the driver on these nodes. I also set debugging on, on one of these for now, so I should get a useful core from this one.

We had a meeting with Tony about the Internet Simulator. I solution to the error messages was found in that the simulator appears to be happy to use incomplete memory maps, if these are accidentally created by aborting the program at a crucial time. The error condition was repeatable so Tony wants a tarball of the damaged scenario, so he can produce a modified version of the simulator where this does not happen. I have run the simulator using the original tarball without any aborted runs, and using my few modifications. This analysis is still running.

04

Jun

2013

Finished fixing the URLs in amp-web so that they are ordered sensibly and can support NNTSC streams that are defined using more than just "source" and "target". I also changed the ordering of the timestamps in the URL so that we can specify a start and end time for the detailed graph only (sensible defaults for the summary graph are meant to be chosen in this case). This is really handy when creating URLs that link to graphs showing events.

Started looking into what needed to be done to prepare NNTSC and netevmon for packaging and a possible distribution for our friends at Lightwire. Spent a decent chunk of time writing a README that should describe exactly how to get a NNTSC instance up and running.

NNTSC and netevmon both have tracs now and I've added a series of tickets to each with the aim of getting a release ready for Lightwire by the end of the month.

03

Jun

2013

Not much happened with the project itself this week as time was spent finishing off the interim report.

02

Jun

2013

Discussed the AMP throughput test with Brendon and decided it was best to move the control messages to a separate socket to the actual test. Now the control socket is used to send the socket options for the test socket before connecting it. So MSS, buffer sizes etc. will be set by both the client and server for the test socket. Discussed CPU and NIC requirements with Brad for the AMP boxes hardware.

31

May

2013

The PHY and channel objects are pretty much done and for the last little while I have been throughput testing and debugging. Initially using a 1.155 Gbps PHY I was reaching only 33 Mbps of measured throughput with a 1500 octet TCP bulk sending application. Moving to UDP speed things up to 100 or so Mbps. Converting the transmit duration calculations from microseconds to nanoseconds speed things up again. After that it was a lot of head banging to realise that the single line of throughput reporting was outputting in Mebibits instead of Megabits. Also getting the exact time in nanoseconds of the UDP flow from when the first packet is received to the last helped bring in the reported throughput to less than 1% difference from the theoretical throughput based on the simulations average number of backoff slots. This difference should be accounted for due to beacon transmissions.

31

May

2013

Handed in my interim report this week.
Also after reading a bit deeper into the changes between openflow 1.0 and 1.2 realised the original approach I was thinking of will not work. So spent some time over the weekend and early this week thinking of other approaches on how I can get a basic spanning tree working in openflow so that when the break comes I'm ready to get right into it.

28

May

2013

This week I've spent most of my time doing my report, plus starting to
rewrite more of the code to make it faster. Also being working on my 560
assignment, building a deep packet inspector in a FPGA which is going
well, hopefully have it running by next week.