Are there any known memory/interrupt issues with the latest version of AIJ

[ghawkins]

I’ve been using AIJ successfully for months, processing tens of thousands of results without any significant problems. However, two days ago, things suddenly started going wrong.

Windows Task Manager now shows AIJ consuming 97% of memory usage, and interrupts have skyrocketed. I’ve tried uninstalling and reinstalling AIJ, but unfortunately, that hasn’t resolved the issue.

As my next step, I plan to roll back to version 5.3.3 to see if that helps. In the meantime, I wanted to ask:

Are there any known issues that might cause this behavior?
Do you have any suggestions or troubleshooting steps I should try?
Please let me know if there’s any specific information or logs I should provide to assist with diagnosing the issue.

Looking forward to your guidance.

Best regards and Happy New Year

Gary

[ghawkins]

So, the rollback did not solve the problem.

Now, it appears to be a memory usage problem. I'm watching AIJ process debayered data sets, and the memory usage continues to increase until the set Maximum Memory limit is reached. At this point, unsurprisingly, AIJ no longer responds. Do you have any idea why this has started to happen suddenly?

Best regards,

Gary

[ghawkins]

I think I'm closing in on the problem.

The memory-related issue occurs while processing over 700 raw science files. These files are debayered from 3008 x 3008 tri-color raw data, resulting in 1504 x 1504 color images, each slightly over 4MB in size.

When processing the TG files using a variable aperture based on the T1 radial profile, memory usage starts at around 31%, spikes to 46%, and then drops back repeatedly. This behavior is consistent and does not cause any crashes.

However, when processing the TB files with the same variable aperture method, memory usage gradually increases but then accelerates rapidly, reaching the 8GB memory limit - at this point, AIJ becomes unresponsive. In contrast, using a fixed aperture allows the TB files to process without issue.

What’s puzzling is that I’ve been using variable apertures successfully for months without encountering this problem.

I’d appreciate any insights or suggestions you might have on how to address this issue.

Best regards,

Gary

[keastrid]

Hi Gary,

Could you please try adding the following to AstroImageJ.cfg (which is next to AstroImageJ.exe)?

-XX:StartFlightRecording=name=Profiling,filename=flight.jfr,settings=profile,loglevel=info

It should generate a file (flight.jfr) in the same folder. If you could reproduce the problem with the options enabled and send us the flight.jfr that would be great.

Kevin

[ghawkins]

I will do that; I'll get back to you shortly.