KBE 3 Report post Posted May 3, 2018 Hello, I am getting the following error after running a calculation. I have done some research and according to that, it's because I dont have enough disk space, but I have 64 GB available in my disk. Can someone please help me out? The calculation took me about 9 hours and I have to restart it. It says that the non linear analysis was ocnverged but it didnt create a .h3d file to view with hyperview. *** INTERNAL PROGRAMMING ERROR # 8008 *** CHECK detects overwriting of data block: plsgau. * DATAMANAGER ERROR in subroutine "datamx". Array "cs2sres" was the first in memory after this array. Thank you Kind regards, KBE Quote Share this post Link to post Share on other sites
Guest Report post Posted May 3, 2018 Hi @KBE Can you share the .out file, please? Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 3, 2018 45 minutes ago, Prakash Pagadala said: Hi @KBE Can you share the .out file, please? Hi Prakash, Unfortunately I can't, as it's confidential. Could you please suggest some things I could verify in my set up? I am also getting the following error with it. *** INTERNAL PROGRAMMING ERROR # 8011 *** DELETE detects overwriting of data block: plsgau. * DATAMANAGER ERROR in subroutine "datamd". Re-checking all lenvec buffers Quote Share this post Link to post Share on other sites
Nachiket Kadu 9 Report post Posted May 3, 2018 36 minutes ago, KBE said: Hi Prakash, Unfortunately I can't, as it's confidential. Could you please suggest some things I could verify in my set up? I am also getting the following error with it. *** INTERNAL PROGRAMMING ERROR # 8011 *** DELETE detects overwriting of data block: plsgau. * DATAMANAGER ERROR in subroutine "datamd". Re-checking all lenvec buffers Hi Under load you have placed Disc deplacement load colletor? Does it contain force data of SPC data? Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 3, 2018 24 minutes ago, Nachiket Kadu said: Hi Under load you have placed Disc deplacement load colletor? Does it contain force data of SPC data? Hi @Nachiket Kadu The disc displacement is made of SPCDs. Quote Share this post Link to post Share on other sites
Nachiket Kadu 9 Report post Posted May 3, 2018 Hi You are trying to give displacement? If yes try giving it in SPC itself and then check. Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 3, 2018 35 minutes ago, Nachiket Kadu said: Hi You are trying to give displacement? If yes try giving it in SPC itself and then check. I have done a model similar to this. The only difference is the size of the two cylinders that I am trying to encase. I have already defined the necessary SPCs and the SPCDs for those SPCs. It worked with the first model but not working with this. I double checked all the constraints as well. Quote Share this post Link to post Share on other sites
Guest Report post Posted May 3, 2018 Hi @KBE Can you share the list of output requested you have in your solver file? Also, remove NLOUT and check if that helps. Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 3, 2018 31 minutes ago, Prakash Pagadala said: Hi @KBE Can you share the list of output requested you have in your solver file? Also, remove NLOUT and check if that helps. Hi @Prakash Pagadala, Is this what you meant? or did you want the output data requested in the global_output_request card? Quote Share this post Link to post Share on other sites
Guest Report post Posted May 3, 2018 27 minutes ago, KBE said: or did you want the output data requested in the global_output_request card? Yes, Some result outputs are not compatible with NLOUTs and may end up using high scratch files. It would be better if you share at least the memory required and available memory information from .out file. Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 3, 2018 28 minutes ago, Prakash Pagadala said: Yes, Some result outputs are not compatible with NLOUTs and may end up using high scratch files. It would be better if you share at least the memory required and available memory information from .out file. Did you want the information below? MEMORY ESTIMATION INFORMATION : ------------------------------- Solver Type is: Sparse-Matrix Solver Direct Method Current Memory (RAM) : 12498 MB (Note: MUMPS Solver Allocates Memory Separately.) Estimated Minimum Memory (RAM) for Out of Core Solution : 15048 MB Recommended Memory (RAM) for Out of Core Solution : 15388 MB Recommended Memory (RAM) for In-Core Solution : 23967 MB (Note: Memory for MUMPS Solver is Included in Estimation.) DISK SPACE ESTIMATION INFORMATION : ----------------------------------- Estimated Disk Space for Output Data Files : 556 MB Estimated Scratch Disk Space for In-Core Solution : 1396 MB Estimated Scratch Disk Space for Out of Core Solution : 9975 MB ************************************************************************ As I am running another calculation right now, I have 46Gb still available in my main disk. Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 3, 2018 36 minutes ago, Prakash Pagadala said: Some result outputs are not compatible with NLOUTs and may end up using high scratch files. @Prakash Pagadala, See below for the global_output_request card details: everything else is not active. Quote Share this post Link to post Share on other sites
Guest Report post Posted May 3, 2018 @KBE clearly, the issue is with the memory and you can see this from the estimated and available memory information, Can you remove GAUSS stress and try again? Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 3, 2018 18 minutes ago, Prakash Pagadala said: Can you remove GAUSS stress and try again? Yeah i could, but it takes hours for the calculation. Should i just run it and send you the memory information? Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 3, 2018 25 minutes ago, Prakash Pagadala said: Can you remove GAUSS stress and try again? Tried this but there still isnt enough memory space. Is there another option? Quote Share this post Link to post Share on other sites
Guest Report post Posted May 4, 2018 Hi, Can you share the output file (memory information from the latest) You can use a temporary directory to let solver write scratch files. Please see the link: And, sorry for a late reply. Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 4, 2018 2 minutes ago, Prakash Pagadala said: You can use a temporary directory to let solver write scratch files. Please see the link: And, sorry for a late reply. Hi Prakash, I tried the first two steps that you mentioned yesterday. Closing all applications running and cleaning the disk, but I still had the same issue. Where can I find the fixlen option? Is it in optistruct options or do I have to define it in a card? Currently, my collegue is using the licence so I have to wait till he is done. I will try what you said and let you know. Thank you KBE Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 4, 2018 Prakash, I tried launching the calculation with fixlen but each time I enter a value for the memory, it requires more and the calculation fails. I first tried with 500, it required 540, so i entered 600 then it required 820 so then i entered 1000 and this is what i got. *** ERROR # 7001 *** Not enough memory for preprocessing (total number of nonzero terms has not been determined) OptiStruct was executed with fixed memory size, which was not enough for the given run. Please allocate more memory or use automatic memory allocation feature in OptiStruct. This error occurs in module "prpdrv". ************************************************************************ RESOURCE USAGE INFORMATION -------------------------- MAXIMUM MEMORY USED 1000 MB MAXIMUM DISK SPACE USED 1 MB ************************************************************************ ************************************************************************ *** ERROR # 7000 *** Not enough memory for completing the run current memory (in MB) = 1000 required memory (in MB) = 1070 OptiStruct was executed with fixed memory size, which was not enough for the given run. Please allocate more memory or use automatic memory allocation feature in OptiStruct. The problem is, I did a similar calculation with the exact same configurations and it worked fine. Regards, KBE Quote Share this post Link to post Share on other sites
Guest Report post Posted May 4, 2018 @KBE Fixlen will not help. You need to use TEMPDIR to write out scratch files. Please try that and if you still face the same error, then I request you to share the .out file to report the issue with developers. Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 4, 2018 I tried with tempdir as well and i have the same issue. Quote Share this post Link to post Share on other sites
Guest Report post Posted May 4, 2018 3 minutes ago, KBE said: Tried with maxlen and this is what i get. ************************************************************************ MEMORY ESTIMATION INFORMATION : ------------------------------- Solver Type is: Sparse-Matrix Solver Direct Method Current Memory (RAM) : 12482 MB (Note: MUMPS Solver Allocates Memory Separately.) Estimated Minimum Memory (RAM) for Out of Core Solution : 15032 MB Recommended Memory (RAM) for Out of Core Solution : 15372 MB Recommended Memory (RAM) for In-Core Solution : 23951 MB (Note: Memory for MUMPS Solver is Included in Estimation.) DISK SPACE ESTIMATION INFORMATION : ----------------------------------- Estimated Disk Space for Output Data Files : 556 MB Estimated Scratch Disk Space for In-Core Solution : 584 MB Estimated Scratch Disk Space for Out of Core Solution : 9163 MB *********************************************************************** I still see the issue with memory, try with -minlen Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 4, 2018 heres what i got after trying with tempdir. It's exactly the same thing. Current Memory (RAM) : 12482 MB (Note: MUMPS Solver Allocates Memory Separately.) Estimated Minimum Memory (RAM) for Out of Core Solution : 15032 MB Recommended Memory (RAM) for Out of Core Solution : 15372 MB Recommended Memory (RAM) for In-Core Solution : 23951 MB (Note: Memory for MUMPS Solver is Included in Estimation.) DISK SPACE ESTIMATION INFORMATION : ----------------------------------- Estimated Disk Space for Output Data Files : 556 MB Estimated Scratch Disk Space for In-Core Solution : 584 MB Estimated Scratch Disk Space for Out of Core Solution : 9163 MB ************************************************************************ For minlen, what value should I put? Quote Share this post Link to post Share on other sites
Guest Report post Posted May 4, 2018 Start with 16000MB Quote Share this post Link to post Share on other sites
KBE 3 Report post Posted May 4, 2018 should I run with out-of-core instead of in-core? Current Memory (RAM) : 16000 MB (Note: MUMPS Solver Allocates Memory Separately.) Estimated Minimum Memory (RAM) for Out of Core Solution : 15032 MB Recommended Memory (RAM) for Out of Core Solution : 15372 MB Recommended Memory (RAM) for In-Core Solution : 23951 MB (Note: Memory for MUMPS Solver is Included in Estimation.) DISK SPACE ESTIMATION INFORMATION : ----------------------------------- Estimated Disk Space for Output Data Files : 556 MB Estimated Scratch Disk Space for In-Core Solution : 584 MB Estimated Scratch Disk Space for Out of Core Solution : 9163 MB Quote Share this post Link to post Share on other sites
Guest Report post Posted May 4, 2018 1 minute ago, KBE said: should I run with out-of-core instead of in-core? Current Memory (RAM) : 16000 MB (Note: MUMPS Solver Allocates Memory Separately.) Estimated Minimum Memory (RAM) for Out of Core Solution : 15032 MB Recommended Memory (RAM) for Out of Core Solution : 15372 MB Recommended Memory (RAM) for In-Core Solution : 23951 MB (Note: Memory for MUMPS Solver is Included in Estimation.) DISK SPACE ESTIMATION INFORMATION : ----------------------------------- Estimated Disk Space for Output Data Files : 556 MB Estimated Scratch Disk Space for In-Core Solution : 584 MB Estimated Scratch Disk Space for Out of Core Solution : 9163 MB Do you still get the error? Quote Share this post Link to post Share on other sites