0 Mbps throughput with UL traffic

[Romieu Ugo]

Hello everyone,

I noticed that when doing uplink traffic simulations with different examples (here traffic-ngmn-mixed), the displayed throughput between the UE and the remote host is equal to 0. Indeed, no packets seem to arrive at the remote host... I'm working in 4.1 version.

If I understand the addressing plan correctly:
1st UE IP: 7.0.0.2 --> SGW IP: 7.0.0.1 --> PGW IP: 1.0.0.1 --> remoteHost IP: 1.0.0.2

In the code:

line 875

auto [remoteHost, remoteHostIpv4Address] =

nrEpcHelper->SetupRemoteHost("100Gb/s", 2500, Seconds(0.000));

line 1785

dlClientLowLat.SetAttribute(

"Remote",

AddressValue(

addressUtils::ConvertToSocketAddress(remoteHostIpv4Address, dlPortLowLat)));

The print flow is between UE and PGW so not on remote host (?)
Flow 1 (7.0.0.2:49153 -> 1.0.0.1:1234) proto UDP
Is "remoteHostIpv4Address" really the IP address of remote host or PGW? like explain in the documentation https://cttc-lena.gitlab.io/nr/html/classns3_1_1_nr_epc_helper.html#a30c10b6ca2d60dc5bed0abfa98fa7eff

With this code, the result is OK. Remote 1.0.0.2 get rx packets:

Ptr<Ipv4> ipv4 = remoteHost->GetObject<Ipv4>();

Ipv4Address addr = ipv4->GetAddress(1,0).GetLocal();

dlClientLowLat.SetAttribute(

"Remote",

AddressValue(addressUtils::ConvertToSocketAddress(addr, dlPortLowLat)));

clientApps.Add(dlClientLowLat.Install(ue));

Thanks in advance for your help.

Ugo Romieu

[Tom Henderson]

I think you identified a bug.  Line 660 of nr-no-backhaul-epc-helper.cc is returning the PGW address.  It probably should return internetIpIfaces.GetAddress(1, 0)).  SetupRemoteHost6 doesn't seem to have this problem.

Ugo Romieu --
You received this message because you are subscribed to the Google Groups "5G-LENA-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 5g-lena-user...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/5g-lena-users/475cf098-5edf-4b7f-819b-c6daf0e76fc3n%40googlegroups.com.

[Gabriel Ferreira]

Hmm, João just commented about this yesterday. He is working on the fix and adding asserts to prevent this kind of regression to happening again.