I’ll have a look real quick and get back to you.
-Dustin
--
You received this message because you are subscribed to the Google Groups "Powder Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to powder-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/CAHUWr4XrHC8oGP2gsvA0smfaGHOmrCucf0zMNm7AL-_MYyVgzQ%40mail.gmail.com.
Re-flashing the firmware and power cycling the radio worked for me. You should be good to go.
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/C498CF95-28D7-44E1-AF61-87FDAD4D0A18%40gmail.com.
This is usually because the MTU of the interface isn’t set high enough, but I checked and that isn’t the case in your experiment so I’m not sure what is happening. Someone with more networking knowledge will have to chime in, but that might not happen until tomorrow given the holiday.
FWIW, if you aren’t seeing lots of underflows and late packets from UHD (represented by U
and L
) when you are running srsenb
, then it shouldn’t be a problem.
-Dustin
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/CAHUWr4X3OS1sq2jaDt%3Dw-r88rv5HuSXbmkRGsYKZw77yfrvP_A%40mail.gmail.com.
Hi Xin Yao,
Is it always the same node that has the issue? Does it actually affect the throughput of the UE with the issue?
-Dustin
--
You received this message because you are subscribed to the Google Groups "Powder Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to powder-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/CAHUWr4VSWcbU47oZbxwRSf%2BQmiSLutuXmsrUaOZiiYEw%3Dt-GXQ%40mail.gmail.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/D81DDF5B-41BB-4963-ABD5-756B4BD69968%40gmail.com.
OK. Let me have a look. Alright if I poke around a bit now?
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/CAHUWr4Xc31RBPXXJrVfCWih%3DNVNA7zAs%3D6HpNc-DK-_Us5t_jQ%40mail.gmail.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/173C68D3-AB80-499D-BA5F-25A66D24BCA4%40gmail.com.
Found the issue…
When you use this profile with multiple UEs you need to change the default UE configs to use different IMSIs and add corresponding records to /etc/srslte/user_db.csv
on the EPC node. I’ve done this for your experiment and things are working fine now.
FWIW, the eNB log showed a rejected connection, which is how I figured this out. In the future, when you run into issues, attaching full text logs (not pictures of parts of them) might help us debug faster.
-Dustin
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/CAHUWr4UnQqdeRJbAyJ0nXZbR-gkdNz598K73StQ9CAXSJF_02Q%40mail.gmail.com.
Hi Xin Yao,
iperf3
should be installed on the image used in that profile. Run in server mode at the UE side to test the downlink or at the EPC side to test the uplink (client mode at the opposite side).
-Dustin
--
You received this message because you are subscribed to the Google Groups "Powder Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to powder-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/CAHUWr4WZt%3D2jm-n%2B1q-yvO3hUNrAQXOjOrY_ZFuf%3DQviNs2uXg%40mail.gmail.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/5D241166-3801-4049-B266-F0CCBBA1397E%40gmail.com.
I run the iperf command and I saw that it needs the hostname for the client
mode side, the command on the server mode side is 'iperf -s' and 'iperf -c
<host>' for client mode side, but 'iperf -c <host>' shows 'missing name for
redirect', do I have the correct command for both sides?
The commands I run to use iperf3
:
# UL
## on the EPC node
iperf3 -s
## on the UE node
iperf3 -c 172.16.0.1
# DL
## on UE node
iperf3 -s
## on EPC node
iperf3 -c {ip of UE on interface that srsue creates (something like 172.16.0.X)}
It defaults to TCP, but you can use UDP if you like.. just look at the documentation for iperf3
to figure out how.
Moreover, the SNR becomes 0.0 even if I only connect one UE. And, if I want
to change the IMSI for UE to connect multiple UEs at the same time, should
I change the k and IMSI, or do I only need to change the IMSI?
Just the IMSI. This should work (I just tested it on your last experiment).
Alright for me to log in and have a look at your current experiment?
-Dustin
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/CAHUWr4UbnRRXgrKqq8RD8Ov6qeM%2BNrvHCEhaGMC%2B9MFX%2B2iUjQ%40mail.gmail.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/FFF95CCE-4BC6-41C9-B718-2E32C344B051%40gmail.com.
OK. Everything is working as expected. FWIW, your user_db.conf
file only had a single entry and both ue.conf
files had the same IMSIs. I’m including a screenshot of the relevant settings changes (note that I changed imei
for the second UE out of habit; I don’t think this is necessary for what you are doing). I’ve also created backup configuration files on each node with dustin
appended to the filenames so you can refer back to them.
In the future, for all issues, please include a link to your running experiment (just grab this from the browser when you are looking at the experiment page). And for issues like these, please attach relevant config and log files for srsepc/srsenb/srsue.
-Dustin
To view this discussion on the web visit https://groups.google.com/d/msgid/powder-users/CAHUWr4XeOXbw_1pXubxf%3DxdLz1nzWbuWZbhqjHBy77DbcTkw8w%40mail.gmail.com.