×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

Memory issues with Kstars/Ekos on Raspberry PI 4

  • Posts: 16
  • Thank you received: 1
Hi Jasem!
Yes, when I was preparing the session on the Crescent yesterday and I ran load and slew to recenter the target from the previous night, it crashed during solving.
This is the FIT file , I attach the log as well in case it could be useful. :)
2 years 8 months ago #73024

Please Log in or Create an account to join the conversation.

That solved in 2.87 seconds in StellarMate and GOTO worked as expected without any issues. So it must be something else that is causing the crash and not memory (have 4GB on SM)

2 years 8 months ago #73029
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 16
  • Thank you received: 1
Thank you Jasem for taking the time to test it!
This doesn't happen for me either if I use the simulator, however it happens "in production" with my gears.
I have the following setup:
- SW EQ6 Pro
- ASI 1600 GT (same as 1600 MM but with EFW incorporated SBIG-like)
- ASI 290 MM mini for guiding
- ASI EAF Motor
- Pegasus Power Box Advance

I run everything on the Raspberry to which I connect using VNC via browser or alternatively TeamViewer if I'm away from home.
Did you also have a look at the log file? Any suggestion on how I can try to get more informations on what's going on? I'll try to run KStars with gdb tonight, let's hope it will give me a more descriptive output of the hard crash.
Anyway thanks again for your time, I really appreciate it!
2 years 8 months ago #73031

Please Log in or Create an account to join the conversation.

  • Posts: 1029
  • Thank you received: 301
There's a huge amount of eqmod serial communication in the logs. If the mount is out of the equation and that level of detail is not needed, it may help the Pi to disable debug logs for that device.

-Eric
2 years 8 months ago #73032

Please Log in or Create an account to join the conversation.

  • Posts: 16
  • Thank you received: 1
Yes indeed Eric, I just enabled all the logs to find the culprit of this, but as I progress with my bug hunt I'll surely clear up the logs from unnecessary data.
Thanks for having a look at it!
2 years 8 months ago #73033

Please Log in or Create an account to join the conversation.

  • Posts: 349
  • Thank you received: 107
Try create new profile with your devices reproduce crash and then remove them one by one. Like EAF, power box ASI290, EQ6.
2 years 8 months ago #73037

Please Log in or Create an account to join the conversation.

  • Posts: 53
  • Thank you received: 8
I have similar gear as you (have a ASI1600MM-Pro not a GT and have a different focuser), I might see if I can try this. I also run Manjaro ARM rather than Astroberry/StellerMate or Ubuntu based, though I could try to burn an image something default. I can also try to run a test from my x86_64 server. I'll let you know how this goes.
The following user(s) said Thank You: Marco Rapino
2 years 8 months ago #73075

Please Log in or Create an account to join the conversation.

  • Posts: 16
  • Thank you received: 1
Wow, thank you very much Will! It's hard to replicate, some nights it happens even 3-4 times, some others, like yesterday, it didn't happen at all. I tried using local astrometry as solver, but it failed several time to solve and didn't perform as well as the internal solver, I'm not sure why.
2 years 8 months ago #73082

Please Log in or Create an account to join the conversation.

  • Posts: 421
  • Thank you received: 102
Odd that you're getting different results with local astrometry vs. internal solver, since to the best of my knowledge, internal solver just took astrometry.net, and turned it into a library. Although the speed does depend on which source extraction method you use. Using "Internal SEP" for the extraction method, and "local astrometry" for the solver, gives great, reliable performance for me. Solves are always under 2 seconds, and often under 1 second. This is on my Raspberry Pi 4 4GB. Maybe you have too many index files downloaded and installed. My understanding is having too many index files can slow it down (but not cause it to be less reliable).

You can also try ASTAP. I like ASTAP, and use it often. I really like the simplicity of index files with ASTAP. It doesn't use the astrometry index files, it has its own set. But you only have to download a single index file that covers everything, not 10 separate files like you do for astrometry.net (and internal solver).
2 years 8 months ago #73112

Please Log in or Create an account to join the conversation.

  • Posts: 349
  • Thank you received: 107
Does it always crash on same file? Try to check md5sum of that file and compare it with this data.astrometry.net/4200/md5sums.txt Just run "md5sum index-4200-07.fits"
2 years 8 months ago #73113

Please Log in or Create an account to join the conversation.

  • Posts: 16
  • Thank you received: 1
Hey Kevin, I actually managed to get the local astrometry solver to work. Since then, at least until now, I haven't experienced any crash whatsoever, so most likely there's an issue with the internal solver.
Will update this post in case I have some update in the future.
Thanks everybody for your help! :)
2 years 8 months ago #73218

Please Log in or Create an account to join the conversation.

  • Posts: 421
  • Thank you received: 102
Excellent, glad to hear it!

I like the idea of the internal solver, especially for other platforms (Windows and Mac) where it's a lot more difficult to install astrometry.net locally. I'm sure they'll get it worked out. But in the meantime, local astrometry.net works just fine, and is plenty fast (for me).
2 years 8 months ago #73219

Please Log in or Create an account to join the conversation.

Time to create page: 2.823 seconds