×

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

Bi-monthly release with minor bug fixes and improvements

Issues with autofocus and meridian flip after a crash [solved]

  • Posts: 1009
  • Thank you received: 133
Last night kstars crashed during an acquisition session :(
I restarted it, fired up EKOS and connected again (it is set up to use web manager on a remote machine).  All devices came up still in proper state, i.e., the CCDs had their cooling point active, the mount had proper coordinates and was tracking.  So I started guiding, and then continued the capture sequence (after setting it up again - this is only capture, not via scheduler).  Everything worked nicely, until:
some 50 minutes later, a filter change was on schedule.  The system is set up to do an AF run on filter change.  That did not happen.  It only applied the specified filter offset, without running AF.  I aborted the capture, and manually triggered the AF.  Then I continued.
On the next filter change, the AF run was done automatically.

Question:  Does the automatic AF on filter change only happen if the current instance of EKOS had done an AF run before?  If so, is this intentional, and why?

Next problem occurred later in the night when the meridian flip was supposed to take place.  It did not happen, although it was activated in the mount tab (6 minutes after crossing).  So instead, the mounts 'safety limit' setting did the flip 20 minutes later.  Of course this aborted the guiding (and thus the whole acquisition).  

Question:  Does the MF only happen if the current instance had done a plate solve before?  Or are there other reasons why MF wasn't done?
 
2 years 11 months ago #70211

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

1. Correct. Sequence job sets policy to autofocus in FilterManager only if prior autofocus was successful.
2. Need log.

If these issues are happening in GIT, then we must resolve them ASAP. The crash + MF issues should are both show-stoppers. I will try to check MF tonight in the observatory.
2 years 11 months ago #70212

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

  • Posts: 1009
  • Thank you received: 133
 
Yes, that one I sort-of remembered, as it also doesn't do timed/T-based AF without an initial run. Does make sense for the T-based one, as that depends on a reference temperature. Not so much for the other two, but I assume making a switch case there would not be worth the complications....
Attached. The first one, log_20-22-33.txt, is until the crash, the second one, log_22-42-04.txt, has the continuation session. Unfortunately only default logging for both, so not too much info I'm afraid.
Yes, that is git, though somewhat old: kstars-3.5.3-1950_gccd592aa2, state as of March 31.
I'm somewhat busy ATM with work, but will try to update kstars today, and have a try with a forced kill and re-connect (if sky permits) tonight (or any other test you'd suggest)
2 years 11 months ago #70214
Attachments:

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

Neither log contains any information about meridian flips. I noticed one log ending with socket errors (which are odd) about X11 server issue.
2 years 11 months ago #70216

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

  • Posts: 1009
  • Thank you received: 133
Yeah, that's what I was afraid of. I would have hoped that there's some mention of disabling MF for whatever reason. Maybe there was (only) something in the mount tab (greyed out option or niotice)? But I hardly ever actually look at the mount tab. I'll have an eye on that tonight.

As for the crash - yes, the error message sounds weird. KStars was running locally on my laptop, so the X server definitely was up, and didn't have any issues. At least there's nothing in any of the system logs...
So likely not an issue of kstars.
2 years 11 months ago #70217

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

  • Posts: 1009
  • Thank you received: 133
So I guess you can lean back.  I checked the MF issue, and as I was afraid, it was user error, sort of.  Even thhough the mount was on target, EKOS was lacking a defined target, and therefore has de-activated MF.  It doesn't need a plate solve, but at least a slew to a position before MF starts being active.
The AF issue had been clarified already.  And the crash probably remains a mystery.  I wanted to have a look at the core file, but it was truncated, being too large.  So no help there either I assume...

(But the upgrade to latest KStars brought some really nice features and fixes :D )
 
2 years 11 months ago #70242

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

Time to create page: 0.575 seconds