ns3 with real application

165 views
Skip to first unread message

Rabia Saleh

unread,
Sep 5, 2023, 12:03:15 PM9/5/23
to ns-3-users
Hi Tom
I have created a MANET using a Docker container and ns3 nodes as a hybrid network as shown in the graph below, but when I increase the number of ns3 nodes to 50 the ping stops working (docker container ping ns3 nodes) any suggestion,manet.png please?

Tom Henderson

unread,
Sep 6, 2023, 11:03:07 AM9/6/23
to ns-3-...@googlegroups.com, Rabia Saleh
What have you tried to do to debug this? I recommend the following
strategy.

1) find the number of nodes for which it stops working. For instance,
if it works for 29 and stops for 30, then you have two configurations to
compare

2) using ns-3 logging or using tools in the real network like tcpdump,
try to find the point at which you do not see the packets anymore
> ns3 nodes) any suggestion,manet.png please?
>
Message has been deleted
Message has been deleted

Tommaso Pecorella

unread,
Sep 6, 2023, 5:58:40 PM9/6/23
to ns-3-users
My best guess is that your computer isn't powerful enough to run that many docker images [fast enough].

The problem is that if the ns-3 VM does not get enough CPU cycles, the ns-3 scheduler won't be able to process the events fast enough to satisfy the real-time requirements, and everything will go... bad.
If this happens, you could increase the priority of the ns-3 VM w.r.t. the other containers. This document could be relevant, but I never did experiment with it: https://docs.docker.com/config/containers/resource_constraints/

Another possibility is that the client and servers (the "native VMs") are filling the ns-3 channel, causing congestion. In that case the system can (basically) stop responding for every machine. A solution to that is to decrease the traffic rate, and to add a tail drop (or a more advanced queue management) to the NetDevices handing the Tap interfaces. This won't help a lot, but at least you'll see *something* passing through.

Tom Henderson

unread,
Sep 6, 2023, 6:28:31 PM9/6/23
to ns-3-...@googlegroups.com
Another common problem when starting a large MANET simulation is that
the routers cause congestion as they simultaneously all try to
synchronize their databases. Try starting them one by one (e.g. start
one routing process, wait 5 seconds, start the next one, etc.).

Rabia Saleh

unread,
Sep 11, 2023, 9:40:19 AM9/11/23
to ns-3-users
Hi Tommason and Tom
I greatly appreciate your reply. In this scenario, I used just one docker container and 49 ns3 nodes, and I think my laptop is good enough to run this scenario as I do not use a VM I use real Ubuntu OS. In other words, just one node is attached to the tap-bridge and docker container. 

Tommaso Pecorella

unread,
Sep 11, 2023, 10:15:07 AM9/11/23
to ns-3-users
I'd have to investigate it. If you want you can contact me privately on Zulip (to debug it I'll need the scripts and all the details to reproduce it on one of my servers).

Rabia Saleh

unread,
Sep 13, 2023, 2:08:05 PM9/13/23
to ns-3-...@googlegroups.com
Daer Tommason
Many thanks for your reply; it is much appreciated.
Yes, please I would like to contact you privately on Zulip. 


--
Posting to this group should follow these guidelines https://www.nsnam.org/wiki/Ns-3-users-guidelines-for-posting
---
You received this message because you are subscribed to a topic in the Google Groups "ns-3-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ns-3-users/v7Il43BQ9d0/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ns-3-users+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ns-3-users/171f1009-52e8-470d-8af7-38deaf01cfc5n%40googlegroups.com.

Tommaso Pecorella

unread,
Sep 13, 2023, 3:25:19 PM9/13/23
to ns-3-users
The link is https://ns-3.zulipchat.com (it's the official ns-3 zulip channel)
Reply all
Reply to author
Forward
0 new messages