Duet rebooting mid print

vrghost

Ugh, I've just noticed the

@vrghost said in Duet rebooting mid print:

reason: OutOfMemory

message.

I guess I've got little bit too carried away with the scripting. Is there any way I can release a variable or get more RAM?

dc42

@vrghost I am surprised that you are running out of memory. Please share the file that you are trying to print.

vrghost

@dc42
Here is the file
Tool-Changer Cover Long.gcode

And here is the event log from the last print I've attempted (with the gcode whereabouts)
eventlog.txt

Phaedrux

Can you also share your config.g and any other macros that get called along the way?

mrenz999

@vrghost I had a similar issue. Did not reboot just stopped printing twice now. Once 3 mins in the second time 6 mins in. I checked on printer both times to find the heaters off, bed and tool, and it stopped. No errors reported just not printing.

vrghost

@phaedrux
Of course I can.

https://github.com/VRGhost/tc-rrf/blob/main/dist/ilo-tc/sys/config.g

All other scripts can be found in that repo too.

gloomyandy

@mrenz999 If your board is not resetting it is likely that you have a different problem, so you will probably get better help if you start a new thread and describe exactly what happens along with posting the output from running m122 along with your config.g file.

mrenz999

@gloomyandy Will do. My heaters just turned off mid print. So no board reset.

dc42

@mrenz999 said in Duet rebooting mid print:

@vrghost I had a similar issue. Did not reboot just stopped printing twice now. Once 3 mins in the second time 6 mins in. I checked on printer both times to find the heaters off, bed and tool, and it stopped. No errors reported just not printing.

Did you run M122 and check the last reset time, to see if it was caused by a power glitch?

vrghost

Hello,

In my attempts to make a minimal code to replicate the issue, I've stumbled on another bit of unexpected behavior.

If one declares a local variable on the top level of a "job" (a file that is put into the "gcodes" directory on the duet), that local variable is NOT released upon completion of the job. So, a second run of the same gcode is aborted with the

M32 "0:/gcodes/min_bug.gcode"
File 0:/gcodes/min_bug.gcode selected for printing
Error: in GCode file line 2 column 16: meta command: variable 'outer_idx' already exists
Cancelled printing file 0:/gcodes/min_bug.gcode, print time was 0h 0m

message.

To replicate just run this gcode file as a print job twice.
min_var_bug.gcode

vrghost

Also, is "abort" command preventing GC of local vars?

The file below works as expected on the first run, but the execution starts with "outer_idx" set to 3 when running the job second time.

min_abort_bug.gcode

dc42

@vrghost said in Duet rebooting mid print:

If one declares a local variable on the top level of a "job" (a file that is put into the "gcodes" directory on the duet), that local variable is NOT released upon completion of the job.

That one has already been reported, and fixed in 3.4beta3.

dc42

@vrghost I've run that macro several times, using M98 from DWC, and it produces the same output each time. So I think this one is also fixed already.

Did you manage to create a file that causes all the memory to be used up?