[Mini-NDN] Issues with NACKs

[Saurab Dulal (sdulal)]

Hi Andre,

First, let's try by increasing sleep time after starting NLSR. Depending upon the size of the topology, NLSR requires enough time for route computation and convergence otherwise it's possible to get many noRoutes nacks.

Additionally, check if the routes (from consumer to the producer) are properly set up or not via Mini-NDN CLI.

If the above step doesn't work, please send here NLSR and NFD debugs log files and the topology that you are using.


Saurab,

________________________________
From: Mini-NDN <mini-ndn-bounces at lists.cs.ucla.edu> on behalf of Andre via Mini-NDN <mini-ndn at lists.cs.ucla.edu>
Sent: Tuesday, February 2, 2021 7:03 PM
To: mini-ndn at lists.cs.ucla.edu <mini-ndn at lists.cs.ucla.edu>
Subject: [Mini-NDN] Issues with NACKs

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and trust the content is safe.


Hello everybody,


I've been experimenting with some custom topologies with a bunch of
producer-consumer pairs. However I've been getting a lot o NACKs when
transmitting data.

These producer/consumer programs are the same as those in the
ndn-cxx/examples directory with some slight modifications only to log
results in a custom format. They are instantiated in a python script,
based on the one in examples/mnndn.py, which sets everything up and runs
MiniNDN.

Here are some details of the issue:

1. I still get data through, though in a small proportion in relation to
the NACKs, and it is not related to any specific links or hosts in the
topology.

2. It is not specific to my setup, which was created using the
Vagrantfile in the last few commits on the named-data/mini-ndn
repository, since I got similar results in different computers.

3. It not a hardware limitation, since I've tried running it with up to
16GB o RAM and 12 CPU cores.

4. It is not caused by the lack of the nlsrc advertise command. In fact,
I'm currently using the default filters pre-advertised in nslrc
(/ndn/<host_name>-site/<host_name>/).

5. It is currently working on my laptop, where I can run the simulations
and get 0 NACKs (as I should). However, running the exact same code and
configuration on another installation, which was set up in the same way,
yields a 10-1 proportion of DATAs-NACKs.


I've tried using debug level logs for NFD and NLSRC, but I could not get
anywhere with those since they are pretty confusing and don't seem to
lead me anywhere (some tips would be appreciated).


My hope is that there is something missing (like a command or
configuration not done by the install.sh script) in my setup or code, or
that there is a specific configuration to the VM that I do not have.


Best regards,

André Dexheimer Carneiro


_______________________________________________
Mini-NDN mailing list
Mini-NDN at lists.cs.ucla.edu
http://www.lists.cs.ucla.edu/mailman/listinfo/mini-ndn
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/mini-ndn/attachments/20210203/ef2ee33f/attachment.html>