×

INDI Library v2.0.7 is Released (01 Apr 2024)

Bi-monthly release with minor bug fixes and improvements

New All Sky camera management software: indi-allsky

  • Posts: 5
  • Thank you received: 0
Hi Aaron,

I'm setting up a second indi-allsky with an ASI 662 MC, unfortunately it reports :
Jun 22 01:01:25 allsky2 [INFO] MainProcess allsky._queueManualTasks() #2432: Checking for manually submitted tasks
Jun 22 01:01:25 allsky2 [INFO] MainProcess indi.getTelescopeRaDec() #840: Telescope Coord: RA 17.74, Dec 44.65
Jun 22 01:01:25 allsky2 [INFO] MainProcess indi.getCcdTemperature() #855: Sensor temperature: 31.0
Jun 22 01:01:24 allsky2 [INFO] MainProcess allsky.detectMoonMode() #1614: Moon altitude: -10:07:54.8, phase 13.1%
Jun 22 01:01:24 allsky2 [INFO] MainProcess allsky.detectNight() #1597: Sun altitude: -21:48:36.7
Jun 22 01:01:24 allsky2 [INFO] MainProcess allsky._startImageWorker() #780: Starting ImageWorker process 14
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception: 
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception: 
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception: cv2.error: OpenCV(4.7.0) /io/opencv/modules/core/src/matrix_iterator.cpp:71: error: (-215:Assertion failed) A.size == arrays[i0]->size in function 'init'
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:     adu = cv2.mean(src=data_mono, mask=self._adu_mask)[0]
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:   File "/home/pi/indi-allsky/indi_allsky/image.py", line 1946, in _calculate_8bit_adu
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:     return self._calculate_8bit_adu(i_ref)
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:   File "/home/pi/indi-allsky/indi_allsky/image.py", line 1932, in calculate_8bit_adu
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:     adu = self.image_processor.calculate_8bit_adu()
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:   File "/home/pi/indi-allsky/indi_allsky/image.py", line 459, in processImage
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:     self.processImage(i_dict)
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:   File "/home/pi/indi-allsky/indi_allsky/image.py", line 258, in saferun
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:     self.saferun()
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception:   File "/home/pi/indi-allsky/indi_allsky/image.py", line 229, in run
Jun 22 01:01:24 allsky2 [ERROR] MainProcess allsky._startImageWorker() #773: Image worker exception: Traceback (most recent call last):
Jun 22 01:01:24 allsky2 [INFO] MainProcess allsky.run() #1048: Exposure state: OK
Jun 22 01:01:24 allsky2 [INFO] MainProcess allsky.run() #1047: Camera last ready: 4.1s

Any hint on this? Pulling from git didn't resolve the issue, rebuinding indi neither.
I even deleted the virtualenv dir to no avail.
Thank you for your help (and your wonderful piece of software...)

MS
10 months 2 weeks ago #93597

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

  • Posts: 268
  • Thank you received: 67

This usually happens when you define a mask where the dimensions of the mask do not match the image generated by the camera.
10 months 2 weeks ago #93599

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

  • Posts: 5
  • Thank you received: 0
Jeeez...you are right. I left the mask of the old camera...which has a different resolution.
Thank you!

MS
10 months 2 weeks ago #93603

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

  • Posts: 268
  • Thank you received: 67

No worries. The error is not obvious. I have updated the code to issue an ERROR log statement if there is a mismatch in the dimensions.
10 months 2 weeks ago #93610

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

  • Posts: 268
  • Thank you received: 67

I fixed the long compile time by locking the cryptography module to a version that has a wheel available for the 32bit ARM platform.

However, the setup.sh script completed without any problems on my Raspberry 3 (32bit bullseye). I made some changes to the python requirements file around the time you posted, maybe try pulling down the latest code to see if it will complete.
10 months 2 weeks ago #93611

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

  • Posts: 83
  • Thank you received: 4
did the git pull and removed the old virtualenvironment folder. I recognized that it still uses the debian11 and not the debian11_32 requirements file. I simply copied debian11_32 to debian11 and it is compiling quickly now. Unfortunately it still shows the same error at the end:

Detected a custom installation of INDI in /usr/local/bin




Distribution: Raspbian
Release: 11
Arch: armv6l
Bits: 32

INDI_DRIVER_PATH: /usr/local/bin
INDISERVER_SERVICE_NAME: indiserver
ALLSKY_SERVICE_NAME: indi-allsky
GUNICORN_SERVICE_NAME: gunicorn-indi-allsky
ALLSKY_ETC: /etc/indi-allsky
HTDOCS_FOLDER: /var/www/html/allsky
DB_FOLDER: /var/lib/indi-allsky
DB_FILE: /var/lib/indi-allsky/indi-allsky.sqlite
INSTALL_INDI: false
HTTP_PORT: 80
HTTPS_PORT: 443





**** Setting up indi-allsky service ****
**** Setting up gunicorn service ****
**** Enabling services ****
**** Setup policy kit permissions ****
**** Ensure user is a member of the systemd-journal group ****
**** Setup rsyslog logging ****
**** Indi-allsky config ****
**** Flask config ****
**** Setup DB ****
./setup.sh: Zeile 1804: 2825 Ungültiger Maschinenbefehl flask db init


The script exited abnormally, please try to run again...

CS, Markus
10 months 1 week ago #93706

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

  • Posts: 83
  • Thank you received: 4
another observation, pyindi-client 1.9.1 gets installed but INDI v2.0.2 is available on the system

Detected INDI version: 2.0.2
Looking in indexes: pypi.org/simple, www.piwheels.org/simple
Collecting pyindi-client
Cloning github.com/indilib/pyindi-client.git (to revision 674706f) to /tmp/pip-install-9or6uhm4/pyindi-client_31d06669c2984bfdbc0a04c442c00160
Running command git clone --filter=blob:none --quiet github.com/indilib/pyindi-client.git /tmp/pip-install-9or6uhm4/pyindi-client_31d06669c2984bfdbc0a04c442c00160
WARNING: Did not find branch or tag '674706f', assuming revision or ref.
Running command git checkout -q 674706f
Resolved github.com/indilib/pyindi-client.git to commit 674706f
Preparing metadata (setup.py) ... done
Building wheels for collected packages: pyindi-client
Building wheel for pyindi-client (setup.py) ... done
Created wheel for pyindi-client: filename=pyindi_client-1.9.1-cp39-cp39-linux_armv6l.whl size=4331103 sha256=e913d2a3206dca534c98e59f648698067e6ad3fcea87183be3e823cc2bb88025
Stored in directory: /tmp/pip-ephem-wheel-cache-hdemezf5/wheels/f5/43/08/a6ab23044641d7a9d7627e6d53b21fbab28ed0731b8ea3dd86
Successfully built pyindi-client
Installing collected packages: pyindi-client
Successfully installed pyindi-client-1.9.1

CS, Markus
10 months 1 week ago #93707

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

  • Posts: 268
  • Thank you received: 67
I am guessing the error is basically that the "flask db init" command is not found. It is acting like the python virtualenv is not activated when the command is being run.

The pyindi-client version reporting is normal. This is the version number that is hard coded in the git repo.
10 months 1 week ago #93710

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

  • Posts: 83
  • Thank you received: 4
manually activating the virtualenvironment and issuing the command flask db init or flask db revision --autogenerate results in an illegal instruction error. I believe it should or does execute app.py with no success. Reson might be a flask version not suited for the old processor in my RPI Zero W. Maybe an older flask version or another wheel would do the trick.

CS, Markus
10 months 1 week ago #93753

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

  • Posts: 268
  • Thank you received: 67
I am reading this is apparently a very common problem with armv6l CPUs like those in the Pi Zero. It is possible it is one of the python wheels installed from a arm7l build and is compiled with CPU instructions that are incompatible with your CPU.

If it is related to one of the python modules and you can figure out which one, recompiling that module directly should fix the problem. It is just a matter of finding the binary with the problem.

Unfortunately, I do not have a Pi Zero to test with. I do have a very old Raspberry Pi 1... it looks like it has the same CPU as the Zero, so I will go run some tests.
10 months 1 week ago #93763

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

  • Posts: 268
  • Thank you received: 67
Believe it or not, I found two modules that were causing the segfault on armv6l: cryptography and bcrypt

You can pull down the latest and code and re-run setup.sh or just take this shortcut:

Activate the virtualenv.
pip3 install cryptography==37.0.4 bcrypt==3.2.2
10 months 6 days ago #93813

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

  • Posts: 83
  • Thank you received: 4
Hi Aaron, thank you very much. It seems to work now!!! Will do further testing and reporting.

CS, Markus
10 months 5 days ago #93823

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

Time to create page: 1.130 seconds