When I try cut files with more than 500k rows, I get fails down or up Z axis (fails down happens more often) and my CNC keeps cutting lower (or higher) the level at which the Z axis failed.
For example, I have the 1500K rows 3D relief G-code on which I failed.
After it I insert after each 250k rows pause on 1 hour (G04 P3600) to be sure that it problem not mechanical (or heat for example) and got same failure result.
After it I split this large file into several small 250K rows each and run one after other without any delay - and I cut everything well and got a positive result.
I think it's a problem to work program with RAM or some thing else. I found in it topic -
G-Code file size limit
post where was write -
Practically it is about 500k lines on P4 2.8Ghz, 1GB ram, geforce 2
And you can turn off the preview in AXIS.
I tried turn off preview in AXIS and switch to other minimal interfaces without backplot, but no get positive effect.
But I have 3Gb ram (1Gb+2Gb) and I think LCNC work with 1Gb RAM at first and after it with second 2Gb RAM.
And when it switch to read from first 1Gb to second 2Gb it get fails.
I think
it is BUG of LinuxCNC
I think it needs to be fixed. But I do not even know where to start looking at the source code to understand what needs to be fixed and how.
The second option is to write a script that would load the file for execution in separate parts, so as not to waste time dividing a large file into parts and not to load each part separately - it is very tedious, especially if you have to wait in the middle of the night when one part is cut to run another
I think if I insert one big RAM memory or two identical, it problem will stop for my case of 1500K rows, but if I load much more rows. when it begin switch again between two RAMs - it problem come up again
So manipulating memory is just running away from problems, not solving them as such.
May be some body can help me, or give me some advice how to solve it problem.
Thank you.
:~$ lscpu
:~$ lshw -short
