Composite size opt insufficient disk space
Hi,
So I was doing a composite optimization, and I'm about to run the second phase, the size optimization. But it cant run, and doesn't completed. It keep stopping and saying insufficient disk space while the computer disk itself is 900 GB free.
The attached .out and .stat file was from the run with optistruct solver directly. Previously I was trying to run the optimization from hypermesh and it just stopped at iteration 0 (like analysis) with no warning, error, and result button.
Thank you, looking forward to any help.
ps. the computer was using altair version 14, and it's a lab computer, so I'm not allowed to add, remove, or update anything in it.
Answers
-
Hi Satyanegara,
This seems to be al memory related issue. I suggest you to use memory control running options (-core out or -maxlen ...).
BR-
Thomas
1 -
The following information may be useful for you:
1 -
Thank you for your reply @Thomas Quenaon and @Q.Nguyen-Dai
I'll try to use running options (memory limit or out-of-core solution) and maybe I'll came back here if anything happened
0 -
I've tried to force in core solutions, and the value of current memory are exactly the same with in core required ram. But it gives the exact same fatal error 8111 in the .out file. I'll try to use out-of-core solutions now.
0 -
Satyanegara said:
I've tried to force in core solutions, and the value of current memory are exactly the same with in core required ram. But it gives the exact same fatal error 8111 in the .out file. I'll try to use out-of-core solutions now.
In core option is the one using only memory, this will require the most of memory.
In descending order of memory usage, you have:
- -core in
- -core out
- -min core
BR-
Thomas
2 -
Thomas Quenaon_21507 said:
In core option is the one using only memory, this will require the most of memory.
In descending order of memory usage, you have:
- -core in
- -core out
- -min core
BR-
Thomas
Thank you sir,
unfortunately my simulation just finished with "end of solver screen output" and "job completed" notifications. But the analysis only done on the iteration 0.
There's no warning or error in the .out file, only NOTE #1405 informing about memory expansion.
0 -
Satyanegara said:
Thank you sir,
unfortunately my simulation just finished with "end of solver screen output" and "job completed" notifications. But the analysis only done on the iteration 0.
There's no warning or error in the .out file, only NOTE #1405 informing about memory expansion.
Can you share the .out File Please ?
0 -
-
Satyanegara said:
Sure,
What option did you choose ? Can you also share .stat file please ?
0 -
Thomas Quenaon_21507 said:
What option did you choose ? Can you also share .stat file please ?
I was using -core out and -uselen
0 -
Satyanegara said:
I was using -core out and -uselen
The run seems to be still on going. Are you sure it stopped ?
0 -
Thomas Quenaon_21507 said:
The run seems to be still on going. Are you sure it stopped ?
To be honest, I don't know, but the solver window are showing me this.
And when I'm closing the solver windows, there's no warning about an ongoing solver run.
0 -
Satyanegara said:
To be honest, I don't know, but the solver window are showing me this.
And when I'm closing the solver windows, there's no warning about an ongoing solver run.
Can you try again withou -uselen and tell me if this works ?
BR-
Thomas
0 -
Thomas Quenaon_21507 said:
Can you try again withou -uselen and tell me if this works ?
BR-
Thomas
Sure, thank you for your help so far sir.
I'll let you know when the computer is accessible for me.
0 -
Thomas Quenaon_21507 said:
Can you try again withou -uselen and tell me if this works ?
BR-
Thomas
Unfortunately the problem is still there sir, with the same NOTE #1405. The run just stopped with "job completed".
0 -
Satyanegara said:
Unfortunately the problem is still there sir, with the same NOTE #1405. The run just stopped with "job completed".
I just found out I could run a composite size optimization using another model with no problem.
But if it's the model that have a problem, there's no warning or error about it.
0 -
Satyanegara said:
I just found out I could run a composite size optimization using another model with no problem.
But if it's the model that have a problem, there's no warning or error about it.
Yes, this is related to the lack of memory. If the second model is smaller it can run.
So you will need to use -min core option or reduce the size of your model.
BR-
Thomas
0 -
I can do a test for you if you can share your input. We have computing server with alot of memory here.
0 -
Thomas Quenaon_21507 said:
Yes, this is related to the lack of memory. If the second model is smaller it can run.
So you will need to use -min core option or reduce the size of your model.
BR-
Thomas
Thank you for your advice sir, I'll try -min core option when the computer is accessible again.
And the model is actually very similar, the only differences are the ply orientation. That's what confuses me a lot.
0 -
Q.Nguyen-Dai said:
I can do a test for you if you can share your input. We have computing server with alot of memory here.
I'd like to, I'll send the HM file when the computer is accessible again.
0 -
Q.Nguyen-Dai said:
I can do a test for you if you can share your input. We have computing server with alot of memory here.
I'm sorry but I can't attach the file, the size is more than 5 MB
0 -
0
-
Thomas Quenaon_21507 said:
Yes, this is related to the lack of memory. If the second model is smaller it can run.
So you will need to use -min core option or reduce the size of your model.
BR-
Thomas
the problem was probably from the model, because it's still giving the same result. Even when I'm using -min core option
0 -
Q.Nguyen-Dai said:
:-(
I'll share my file link in the drive then.
0 -
Q.Nguyen-Dai said:
:-(
Okay, here's the .fem file.
I was using -core in and -maxlen 12600, then change it to -core out and then -core min because I thought it was due to memory issues.
0 -
@Satyanegara : it's terminated!
Here's "hwsolver.mesg":
OPTIMIZATION HAS CONVERGED. FEASIBLE DESIGN (ALL CONSTRAINTS SATISFIED).
A lot of results files. I can not upload here for you. If you want, contact me at nguyendaiquy_AT_gmail.com to know how to download them.
0 -
Q.Nguyen-Dai said:
@Satyanegara : it's terminated!
Here's "hwsolver.mesg":
OPTIMIZATION HAS CONVERGED. FEASIBLE DESIGN (ALL CONSTRAINTS SATISFIED).
A lot of results files. I can not upload here for you. If you want, contact me at nguyendaiquy_AT_gmail.com to know how to download them.
Thank you, I'll contact you right away.
0 -
Satyanegara said:
Thank you, I'll contact you right away.
I used Optistruct version 2020 for this test.
0 -
Q.Nguyen-Dai said:
I used Optistruct version 2020 for this test.
Well, maybe I'll have to look for another pc with an updated version. Thankyou for the information.
0 -
Satyanegara said:
Well, maybe I'll have to look for another pc with an updated version. Thankyou for the information.
I run with "-nt 4"
0