Bad allocation error catch.

[jrjdavidson]

Kia ora,

we've setup a way to automatically setup network tasks, using the agisoft scripts on github as inspiration. It works great, however every now and then a project gets stuck on a bad allocation error loop. Trawling the forum indicates this is a RAM issue. I've tried all the tips I've found on here but haven't been able to solve it so far.

I have a couple of questions:
1) can you confirm that the bad allocation error is always RAM issue? Apparently our machine does not max out the RAM when the issue occurs.
2) is there a way to set up an exception catch for errors like this? I thought we could split the chunk if this type of error was thrown, if only we write a try/except pattern for this type of error.

Thank you!

[Alexey Pasumansky]

Hello jrjdavidson,

In most cases "bad allocation" means insufficient RAM (could be also insufficient free RAM to allocate certain amount of memory).

I think you can use try/except approach for such problems, providing that there is no system service that terminates the process which tries to allocate RAM above certain threshold.

[jrjdavidson]

ok, thanks.

For a network task, how would i try to catch an error like this? should i create a something like

Code: [Select]

#previous code
task = Metashape.Tasks.BuildPointCloud()
tasks.append(task)

#new code
task = Metashape.Tasks.RunScript()
task.code='''
import Metashape
doc = Metashape.Document()
doc.open("path\to\project.psx", ignore_lock = True)
chunk = doc.chunk
try:
    chunk.buildPointCloud()
except Metashape.BadAllocationError: #just guessing how to catch errors
    #code to split chunks here
'''
tasks.append(task)