N310 confirmed. Linux native.
tx_samples_from_file produces underruns. So does a customized version of tx_samples_from_file that uses multiple buffers and threads to guarantee constant send pressure. I understand that the N310 should not do this. I agree that the most likely culprit is the host computer, coupled with inadequate buffering in the N310. Does anyone know how much buffering it provides and what the associated timing is? That is, what is maximum lag between UDP packets before the N310 will experience an underrun condition? In the Ubuntu host system log, there are messages during period in question from NetworkManager. Nothing indicating a connection problem, but consistent with my general believe that NetworkManager is an evil abomination whose primary job is to create trouble and frustration, we will remove it from the workstation and try again. Thanks! Jim At 03:32 PM 4/13/2023, Marcus D. Leech wrote: >On 13/04/2023 17:02, Jim Schatzman wrote: >>Even with all the recommended settings, and a very fast computer that is >>doing nothing except sending the data, it is maybe 50/50 that a 2 hour >>simulation can be conducted without an underrun. The longest run I have been >>able to do without an underrun is about 2.5 hours. >> >>The sample rate is 12.5 Msamp/sec at 16 bit I + 16 bit Q or 400 Mbit/sec. >> >>For our application, that is unacceptable. I need to be able to run for days >>without data loss. >> >>It is a mystery to me why a 10 Gbit connection cannot support 400 Mbit/sec >>UDP reliably. >> >>Any ideas about how we can completely eliminate underruns? >> >>At the moment, I am uncertain whether the problem is occurring on the host or >>on the radio. I suspect the radio, but I will do some testing of the host to >>see what UDP data rate it can support without loss. >> >>Thanks! >With respect, it's unlikely to be the radio, per se (unless something is >BROKEN). Streaming is handled entirely by the > FPGA machinery, and unlike a general-purpose computer, running a >general-purpose OS, performance is utterly > deterministic. It doesn't have "other things" going on that might preempt >what it's doing, etc. > >I think you'd mentioned that the hardware involved is a USRP N310?   While >it does have a built-in LInux machine, since you're > using it from another host, the Linux machine is basically uninvolved in >streaming, other than session set-up and tear-down > via MPM. > >Check the logs on your host machine--does the network interface go up/down >during your streaming? If so, that could > indicate a wiring problem. > >Are you running in a native hardware machine, or in a VM?  Windows or Linux? > >If you simply use the "tx_waveforms" example at your desired sample rate, do >you eventually get underruns or not? > >_______________________________________________ USRP-users mailing list -- >usrp-users@lists.ettus.com To unsubscribe send an email to >usrp-users-le...@lists.ettus.com </x-flowed> _______________________________________________ USRP-users mailing list -- usrp-users@lists.ettus.com To unsubscribe send an email to usrp-users-le...@lists.ettus.com
