[Ground-station] GNU Radio Help Needed: OreSat Team (time sensitive!)

[Roland Turner]

Glenn,

I am absolutely not a GR expert, but I wonder whether you aren't missing 
opportunities to materially simplify the graph that you're asking GR to 
plumb, doing which might allow to avoid the problem if not properly fix it:

  * Are you able to split the TX and RX paths into separate processes?
    This is unsatisfying, but the less complicated a task you're asking
    GR to do, the fewer weird things are likely to happen.
  * Are you able to remove *all* of the not strictly necessary blocks?
    e.g. you've not mentioned a Lime device in your setup, but there are
    LimeSuite blocks. Presumably they're not being scheduled any CPU
    time if the device is not physically present, but why rely on that
    assumption when it's in your power to remove it?
  * Does removing file and GUI sinks avoid the problem? If you, how many
    of those can you live without?
  * Etc.

- Roland


------------------------------------------------------------------------


On 4/1/24 16:25, Glenn LeBrasseur via Ground-Station wrote:
> Michelle,
>
> This is the flowgraph in question:
> https://github.com/uniclogs/uniclogs-sdr/blob/maint-3.10/flowgraphs/OreSat0.grc
>
> We have a different repo organization for our ground station work.
>
> I can also send our testing notes if you like.
>
> Thanks so much,
>
> 73,
> Glenn
>
> On Wed, 3 Jan 2024 at 13:28, Michelle Thompson via Ground-Station 
> <ground-station at lists.openresearch.institute> wrote:
>
>     Greetings all,
>
>     Please read the message below from our friends at OreSat and speak
>     up if you can work with them to diagnose the problem preventing a
>     clean handoff on 11 January.
>
>     OreSat's public repository is here: https://github.com/oresat
>
>     OreSat's 0.5 website is here:
>     https://www.oresat.org/satellites/oresat0-5
>
>     There are quite a few repositories in GitHub. Glenn, would you
>     please point out where the problematic flowgraphs are?
>
>     -Michelle
>
>     -=-=-=-=-=-=-=-=-=-=-=-=-=-=-
>     Glenn LeBrasseur <gleb2 at pdx.edu>
>     Tue, Jan 2, 3:16 PM (21 hours ago)
>     to me, Andrew
>
>     Hey Michelle,
>
>     If you, or you have anyone who you know who is knowledgeable with
>     GNU Radio and is willing to help us with an odd buffering problem,
>     we would be totally appreciative.
>
>     Our team here at OreSat is having a bit of an existential crisis
>     regarding GNU Radio. We are using a HackRF One for transmit and a
>     RTL-SDR for receive, to send our fairly simple GFSK traffic
>     through our ground station, and we are having packet buffering
>     problems. The flow graph communicating through UDP sockets is
>     behaving strangely when sending the packets. At first things were
>     fine when our packets were small, but now varying the packet size,
>     ever so slightly, is causing things to go wrong.
>
>     We are totally willing to share our flowgraphs and testing
>     procedure, as you know we are totally open source, so they
>     currently are online at our GitHub site. We will tell you
>     everything we are doing!
>
>     Anything you could help with would be great, particularly since we
>     are doing integration testing right now of our OreSat 0.5 cubesat,
>     for a handoff on 11-Jan-2024, and this has just cropped up.
>
>     Please let me or Andrew know what's possible.
>
>     Kind regards and 73,
>
>     Glenn, KJ7SU
>     gleb2 at pdx.edu
>
>     Glenn N. LeBrasseur, P.E.
>     glennl at pdx.edu
>     -=-=-=-=-=-=-=-=-=-=-=-=-=-=-
>
>
>
> -- 
> Glenn Noel LeBrasseur
> glenn.lebrasseur at gmail.com
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openresearch.institute/pipermail/ground-station-openresearch.institute/attachments/20240104/b85b77bd/attachment-0001.htm>