Foam fatal io error: cannot find file
Hi, I'm new to using openFoam. I have completed all the tutorials without a problem, however now that I am leading towards my own simulation cases, I wanted to test out the snappyHexMesh. I found a tutorial on snappywiki for a simple cube. I downloaded the case file and created a new new directory for it. Both blockMesh and the snappyHexMesh ran without problem. All looked fine in paraView etc.
I then tried to run simpleFoam and got the following error: --> FOAM FATAL IO ERROR: cannot find file file: /home/adam/OpenFOAM/adam-2.3.0/run/cube/system/system/controlDict at line 0. From function regIOobject::readStream() in file db/regIOobject/regIOobjectRead.C at line 73. FOAM exiting I'd appreciate any help. Thanks, Adam |
Greetings Adam and welcome to the forum!
If you run this command: Code:
pwd Code:
/home/adam/OpenFOAM/adam-2.3.0/run/cube/system/ Code:
cd .. Bruno |
Hi Bruno, thanks for your help however when I run pwd I get the following output:
/home/adam/OpenFOAM/adam-2.3.0/run/cube/system so I'm not able to move back a folder. Apologies if there is something really simple that I'm missing Thanks, Adam |
Hi Adam,
Sorry, I made a bad copy-paste... I should have removed the second "system" from the path on the 2nd code block. I've edited my post and fixed it. The idea is that the case folders look something like this:
Best regards, Bruno |
I had tried that but I get a similar error:
--> FOAM FATAL IO ERROR: cannot find file file: /home/adam/OpenFOAM/adam-2.3.0/run/cube/1/p at line 0. From function regIOobject::readStream() in file db/regIOobject/regIOobjectRead.C at line 73. FOAM exiting |
Hi Adam,
I've got the feeling that you've run snappyHexMesh without the "-overwrite" option. This will create additional time/iteration folders that only have a mesh. If this is the case, I suggest that you run the following steps:
edit: And also have a look at the courses available here: http://openfoamwiki.net/index.php/Ha...enFOAM_courses Best regards, Bruno |
Hi Bruno,
sorry that was my mistake, I'm using openFoam on a virtual window atm and I haven't shared the clipboard so I'm having to type out the error. I did use snappyHexMesh -overwrite but the error should have been: --> FOAM FATAL IO ERROR: cannot find file file: /home/adam/OpenFOAM/adam-2.3.0/run/cube/0/p at line 0. From function regIOobject::readStream() in file db/regIOobject/regIOobjectRead.C at line 73. I will look into those courses, but I could do with getting this running so that I can progress with some of my own simulations that I will base around it. thanks, Adam |
Hi Adam,
Then I need to know more about the case you're trying to run, because it doesn't look like one of the tutorials from OpenFOAM. Don't forget that I can't see what you're seeing... ;) Best regards, Bruno |
Hi Bruno,
edit: I have fixed it now. Thanks for your help Adam |
Dear Adam
How did you solve it. I am also having the same error. I am supercomputing for air flow in a pipe tee geometry. On coarse meshing it was working fine but on dense mesh, I have got this for many processors. I have taken message for only one processor. After reaching 0.10 seconds, I have got error file showing message below: --> FOAM FATAL IO ERROR: [77] cannot find file file: /scratch/director674/shaqqi/tee/LES/75_original/processor77/0.101057/p at line 0. [77] [77] From function regIOobject::readStream() [77] in file db/regIOobject/regIOobjectRead.C at line 73. [77] FOAM parallel run exiting I have tried to delete earlier steps and rerun but to no avail. For the last 3 weeks this problem is stuck, I shall be grateful for your suggestion. Best Regards |
Greetings sihaqqi,
It's very much possible that the error you're getting is due to a different reason from what happened in Adam's case (which I never got to know any details about it :)). Therefore, please describe the steps you've taken to reach that error message. Best regards, Bruno |
Dear Bruno
Greetings on my behalf. Many thanks for your response.I have got this pipe tee geometry built and meshed in Ansys as my work relates to flow induced vibrations. CFD loads shall be imposed on pipe structure. Due to licensing issues for supercomputing as one node on machine requires one license, and due to the number of researchers using Ansys, it was not possible to have many licenses so I opted to run in OpenFoam. I converted my mesh using fluent3DMeshtoFoam. I decomposed using decomposePar and ran it. The geometry is a vertical pipe is connected to another horizontal in the middle to form a tee. I am running LES using pisoFoam. I have adopted the same procedure as on another geometry which was two pipes connected to each other at 90 degree, having a mitre joint. There was no problem in that geometry, however in this geometry I have started running and after reaching 0.1 from 1e-7, it is giving me this error. I tried to remove earlier time steps, did ./Allrun in the beginning but it is just not working. My fluid is air at 75m/s so it is an incompressible solver. Mach is 0.22. Please advise what other information you would like to have? Best Regards Imad |
If I understood you correctly, the solver was running without problems in parallel, running continuously since 1e-7 seconds up until it crashed.
Quote:
|
Hello Bruno
Thanks for your response. My answers are 1. I have tried to restart simulations two ways: one by removing some of the last time steps but it again crashed. 2. OpenFoam 2.1.1 4. Time steps are written after every 2000 steps 5. My space is 32GB remaining so it should be alright. 3. Error file is too big to attach, however a some part I am giving below for you to have a look, this comment that it cannot find the file is throughout. I cannot understand the meaning of this error. It has written all previous time steps successfully. Code:
[1026] Thanks for your time again. Regards Imad |
Hi Imad,
I wanted to see what is the output before the first occurrence of: Code:
FOAM FATAL IO ERROR: Code:
quota If I'm not mistaken, the problem is due to a bug that is present in OpenFOAM 2.1.1. If it would be possible to upgrade to at least 2.1.x, this problem should no longer occur. I say this because you're not the first one to trigger a similar issue. Sometime ago there was someone else with the same or at least a similar problem... I'm not sure how he managed to solve the problem... and I can't find the thread I'm thinking of :(. My final question is this: if you start the case from 0.0s and let it run without stopping, does it still crash at 0.101057s with the same error? Best regards, Bruno |
Hi Bruno
Before FOAM FATAL IO ERROR: there is nothing however I have copied somethings for you from the log file as shown below this para. After this there is nothing. I shall check with system administrator about what you told me about memory. Also I shall ask him to upgrade however it is beyond my control if they would cooperate because our cluster is not university owned. It is ivec in Western Australia which provides supercomputing facilities to 5 technical universities in Western Australia as it is a joint venture. Still I will pursue. If I delete all my files and these again crash it would take at least 10 days because I am using 100 nodes = 1200 CPUs because of LES. Geometry is very big. Horizontal pipe is 10m and vertical pipe is 3m and I have to capture yPlus value of 1. Code:
calculating averages Imad |
Hi Imad,
I've made a test just now, using the tutorial case "incompressible/pisoFoam/les/pitzDaily" and using OpenFOAM 2.1.1. It doesn't seem to be a problem with OpenFOAM 2.1.1, because it was possible to continue the simulation in parallel more than once. I hope you haven't erased the files. Check the contents of one of the folders where the solver complained about the missing "p" file. For example, by running: Code:
ls -l processor1026/0.101057 Check the previous processor and time folders to check if they have the field files "U" and "p" and so on. Another question is: are you using any function objects in "system/controlDict"? And if so, how are they configured to run? Best regards, Bruno |
Dear Bruno
No such time step as 0.101057 exists anywhere. Last time step I have in all processors is 0.1010569. It has 1 folder "uniform" which has "field averaging properties" and "time" files. Aside from this folder, there are 14 more files which are Code:
yPlus Code:
application pisoFoam; Imad |
Quote:
Adam |
Hi Adam and Imad,
@Adam: Thanks for the feedback :) @Imad: Sorry, I wasn't able to answer sooner. I have a feeling I have seen this problem before, but I'm not able to figure out what the exact problem is. But I've managed to deduce that it shouldn't be a problem with disk space quota, because you have this option: Code:
purgeWrite 10; I think that if I could see a good list of the files present in the "processor0" folder, along with the full output of the last attempt of running the case from the latest time snapshot, I might be able to better deduce what the problem is. You can get a good list of the contents of the folder "processor0" if you run this command from within the main case folder: Code:
find processor0 > list_of_files_in_processor0.txt Best regards, Bruno |
All times are GMT -4. The time now is 10:50. |