×

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

Bi-monthly release with minor bug fixes and improvements

KStars memory leak and crash

is this on nightly? Please try to see if you get the crash on nightly.
1 year 8 months ago #84165
The topic has been locked.
  • Posts: 39
  • Thank you received: 9
I also had different BTs from different sessions

This is a segmentation fault
Thread 21 "Thread (pooled)" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x98a0c040 (LWP 21872)]
tcache_get (tc_idx=7) at malloc.c:2934
2934	malloc.c: File o directory non esistente.
(gdb) bt
#0  tcache_get (tc_idx=7) at malloc.c:2934
#1  __GI___libc_malloc (bytes=64) at malloc.c:3042
#2  0xb56fa3bc in SEP::Extract::convert_to_catalog(SEP::objliststruct*, int*, SEP::sep_catalog*, int, int) ()
   from /usr/lib/arm-linux-gnueabihf/libstellarsolver.so.2
#3  0xb56fbce0 in SEP::Extract::sep_extract(SEP::sep_image*, float, int, int, float*, int, int, int, int, double, int, double, SEP::sep_catalog**) ()
   from /usr/lib/arm-linux-gnueabihf/libstellarsolver.so.2
#4  0xb56c41d4 in InternalExtractorSolver::extractPartition(InternalExtractorSolver::ImageParams const&) ()
   from /usr/lib/arm-linux-gnueabihf/libstellarsolver.so.2
#5  0xb56cbb70 in non-virtual thunk to QtConcurrent::RunFunctionTask<QList<FITSImage::Star> >::run() ()
   from /usr/lib/arm-linux-gnueabihf/libstellarsolver.so.2
#6  0xb425ef30 in ?? () from /usr/lib/arm-linux-gnueabihf/libQt5Core.so.5
#7  0xb4267b58 in ?? () from /usr/lib/arm-linux-gnueabihf/libQt5Core.so.5
#8  0xb5797494 in start_thread (arg=0x98a0c040) at pthread_create.c:486
#9  0xb398b568 in ?? () at ../sysdeps/unix/sysv/linux/arm/clone.S:73 from /lib/arm-linux-gnueabihf/libc.so.6
Backtrace stopped: previous frame identical to this frame (corrupt stack?)
(gdb)

This one is a bus error: in this case only guiding was active; previous one a complete scheduled session (focus, align, ecc) was running.
Thread 21 "Thread (pooled)" received signal SIGBUS, Bus error.
[Switching to Thread 0x98a0c040 (LWP 4602)]
0xb56c2028 in ?? () from /usr/lib/arm-linux-gnueabihf/libstellarsolver.so.2
(gdb) bt
#0  0xb56c2028 in ?? () from /usr/lib/arm-linux-gnueabihf/libstellarsolver.so.2
#1  0xb56c46dc in InternalExtractorSolver::extractPartition(InternalExtractorSolver::ImageParams const&) ()
   from /usr/lib/arm-linux-gnueabihf/libstellarsolver.so.2
Backtrace stopped: Cannot access memory at address 0x4f1a51f8
1 year 8 months ago #84166
The topic has been locked.
  • Posts: 349
  • Thank you received: 107

Replied by nou on topic KStars memory leak and crash

It is with quite recent build but with older version of stellarsolver. Will try with latest build of stellarsolver and kstars.
1 year 8 months ago #84174
The topic has been locked.
  • Posts: 9
  • Thank you received: 3
Hi,
I think I have the same issue as here. Crashes every night when scheduling. Doesn't crash doing lots of dark/flat frames without using the scheduler. It's only started crashing after I updates about 3 weeks ago, and hadn't updated previously since Feb 2022. That'll teach me that if it 'aint broke, don't fix it.
Has anyone had any progress on this since the last post over 3 weeks ago ?. I see there is a new indi 1.9.7 released a few days ago. Any fixes in this for this issue ?
Here's hoping :-)
1 year 7 months ago #84922
The topic has been locked.
  • Posts: 969
  • Thank you received: 94
'Crashes every night when scheduling.'

Hi
No logs, so can only guess.
Turning off partitioning in stellarsolver works for some systems.
Or you can use ASTAP for solving and PHD2 for guiding.
Cheers and HTH
Last edit: 1 year 7 months ago by alacant.
1 year 7 months ago #84924
The topic has been locked.
  • Posts: 9
  • Thank you received: 3
Thanks for the quick reply. Looked at logs and it's definitely an out of memory issue :-

[2022-08-01T03:48:48.746 CEST DEBG ][ org.kde.kstars.indi] - Image received. Mode: "Normal" Size: 40368960
[2022-08-01T03:48:48.939 CEST INFO ][ org.kde.kstars.indi] - "FITS" file saved to "/media/astroberry/4311-BCAE/EkosImages/East Veil Nebula/East_Veil_Nebula/Light/East_Veil_Nebula_Light_60_secs_2022-08-01T03-48-48_231.fits"
[2022-08-01T03:48:48.939 CEST DEBG ][ org.kde.kstars.fits] - Reading file buffer ( "38.5 MiB" )
[2022-08-01T03:48:49.703 CEST CRIT ][ org.kde.kstars.fits] - Debayed memory allocation failure. Required Memory: "115.5 MiB" Available system memory: 1.8101e+09
[2022-08-01T03:48:49.772 CEST INFO ][ org.kde.kstars.ekos.capture] - "Download Time: 1.92 s, New Download Time Estimate: 2.23 s."
[2022-08-01T03:48:49.779 CEST INFO ][ org.kde.kstars.ekos.capture] - "Received image 231 out of 290."
[2022-08-01T03:48:49.783 CEST INFO ][ org.kde.kstars.ekos.capture] - "Captured /media/astroberry/4311-BCAE/EkosImages/East Veil Nebula/East_Veil_Nebula/Light/East_Veil_Nebula_Light_60_secs_2022-08-01T03-48-48_231.fits"

I do use stellarsolver and PHD2 for guiding. I'll try your suggestions as soon as I can get some sky to work with.

hairbear.
1 year 7 months ago #84925
The topic has been locked.
  • Posts: 349
  • Thank you received: 107

Replied by nou on topic KStars memory leak and crash

After I upgraded stellarsolver crashing stopped for me.
1 year 7 months ago #84934
The topic has been locked.
  • Posts: 123
  • Thank you received: 13

Replied by Outta on topic KStars memory leak and crash

Hi,


My crashings are unrelated to solver and guiding as I am using PHD2 and Online Astrometry. But I do not have any critical details in log.

You do have, it says that there is no memory available, is you RAM full?
1 year 7 months ago #84974
The topic has been locked.
  • Posts: 9
  • Thank you received: 3
Hi,
I'm using a RPi 4b with 4G. My trace looks like an out of memory issue, but the log says

Available system memory: 1.8101e+09

That's a lot of memory :cheer: and could just be Kstars not printing it out properly.
I won't have any decent skies for about a week I think, so I'll try then.
Could I simulate a session ?. Bit of a newbie on Kstars/INDI so may need to look this up if it's possible.
1 year 7 months ago #84977
The topic has been locked.
  • Posts: 57
  • Thank you received: 4
I updated to the latest stable release and am now seeing this problem as well on Ubuntu 20.04LTS. I noticed last night during an imaging session that the desktop was slowing down as the imaging session went on. The main camera download became much slower, and eventually the guider started dropping frames and the desktop became unresponsive. I re-started and watched the Kstars process gradually consume all the memory until it crashed again. Which logs should I enable that would be the best help for diagnosing the cause?
1 year 7 months ago #85021
The topic has been locked.
  • Posts: 9
  • Thank you received: 3
That's exactly the symptoms that I saw with my Rpi 4b. The desktop slowed down more and more, taking longer and longer to respond. Then the out of memory error..
After I implemented an earlier suggestion to turn off all notification sounds, it seemed to stop slowing down.
I currently have the simulator running doing several hours of tracking and capturing. The system is currently losing about 250MB of free memory every hour. That's not enough to explain what I see, which is memory running out after about 3-4 hours of tracking and capturing. I do have notification sounds turned off though, so I'll redo with them on so see if there is a difference.
1 year 7 months ago #85022
The topic has been locked.
  • Posts: 9
  • Thank you received: 3

I'm on StellarSolver 2.3 and Kstars 3.5.9 stable.
I notice that StellarSolver 2.4 was released 6 days ago. Are you on 2.4 ?
1 year 7 months ago #85023
The topic has been locked.
Time to create page: 0.769 seconds