Linuxcnc getting very slow

My pc likes this program.

In terminal type : lscpu

Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 2
On-line CPU(s) list: 0,1
Thread(s) per core: 1
Core(s) per socket: 2
Socket(s): 1
NUMA node(s): 1
Vendor ID: GenuineIntel
CPU family: 6
Model: 23
Model name: Intel(R) Core(TM)2 Duo CPU E7500 @ 2.93GHz
Stepping: 10
CPU MHz: 2933.000
CPU max MHz: 2933.0000
CPU min MHz: 1596.0000
BogoMIPS: 5851.98
Virtualization: VT-x
L1d cache: 32K
L1i cache: 32K
L2 cache: 3072K
NUMA node0 CPU(s): 0,1
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx lm constant_tsc arch_perfmon pebs bts rep_good nopl aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm sse4_1 xsave lahf_lm kaiser tpr_shadow vnmi flexpriority dtherm

Grotius wrote:

for me its doing task at the moment at F500
Wow. I am running on Ethercat... So very fast protocol.

Is your linux 64 bit?
Do you use servo thread or only base thread 15000?

In your program you are at line 3.6 miljion...

i use 64 bit. You also i see, nice work.

I tested to only use base thread tonight. Never did that before. Just followed the basic ethercat config's but that solved not my problem.
I had to tune ethercat steppers for using it with pulse output module instead of pid.
Servo threat was way to slow for that. Just like parport uses the base thread for real time pulses.

I think your program run's stable, but your internetting goes little slower while running such a huge program.
So 3 miljion lines is no problem at all.

Maybe you can increase the speed by using offset's. I see you have the same product 16 times. By using offset's the speed
will increase dramitically. Welcome to the world of subroutines.

The gremlin goes good. It's now making a screw.

Grotius wrote:

In your program you are at line 3.6 miljion...

i use 64 bit. You also i see, nice work.

I tested to only use base thread tonight. Never did that before. Just followed the basic ethercat config's but that solved not my problem.
I had to tune ethercat steppers for using it with pulse output module instead of pid.
Servo threat was way to slow for that. Just like parport uses the base thread for real time pulses.

I think your program run's stable, but your internetting goes little slower while running such a huge program.
So 3 miljion lines is no problem at all.

Maybe you can increase the speed by using offset's. I see you have the same product 16 times. By using offset's the speed
will increase dramitically. Welcome to the world of subroutines.

The gremlin goes good. It's now making a screw.

It can't be done with offsets they are all different sizes.

With smaller program don't have this problem... Only with long programs

Did you try disabling the preview and backplot?
This should make a huge difference with so large a program

Just for info. I did a restart at line 1.9 miljion. That works. If you know it's a long program i can confirm loading takes a while.

PCW wrote:

Did you try disabling the preview and backplot?
This should make a huge difference with so large a program

I have passed this program today, tomorrow or Wednesday i have a new one to pass and try to disable backplot.

The problem is not opening a big program is at the end of so much working hours

I also have such sloweness in my program when the size of it become big. I see such problem at program size biger than 10Mb. Yes, I have old PC.

As for me, I divide my program on the small parts that suits me.
You can use such O-codes: "o<myfile> call" in main program.

What program are you using for generating G-code? Some programs can divide G-code files automatical after making correspondent settings in post processor.

Did you find a solution?
I use linuxcnc to control a 3d printer, and it's not unusual to have prints that last multiple days.
The control software should only load a small portion of the code and dinamically retrieve more code as it runs.
I know that my pc is quite old, but i don't think it's normal that i have to wait 20 minutes to load a program (without backplot) and then having to use a laggy, constantly swapping pc with full ram.




edit: I just discovered that tklinuxcnc GUI actually DOES load the program dinamically. I loaded the program in 10 seconds without any issues.
Can't this behaviour be ported to gmoccapy or axis?