Hello,
I used a newly compiled version of Elmer-8.3 (compiled on ubuntu 14.04 using CMake as instructed here: viewtopic.php?f=2&t=3742) and found that declaring "Maximum Output Level = 40" results in very large output files.
In the new version, right after the assembly of every solver, the string "GetNOFActive: Number of active elements: xxxx" is getting repeatedly printed as many times as the number of elements - making the output files enourmously large.
I have not noticed this with an earlier version of Elmer. A sample output log file is attached.
-kumar
Elmer-8.3 -- Maximum Output Level 40 gives huge output
-
- Posts: 54
- Joined: 17 Jun 2015, 10:04
- Antispam: Yes
-
- Posts: 54
- Joined: 17 Jun 2015, 10:04
- Antispam: Yes
Re: Elmer-8.3 -- Maximum Output Level 40 gives huge output
Noticed that the log file is not uploaded. Here it is.
- Attachments
-
- solver.log
- Solver log file
- (734.17 KiB) Downloaded 357 times
-
- Site Admin
- Posts: 4826
- Joined: 22 Aug 2009, 11:57
- Antispam: Yes
- Location: Espoo, Finland
- Contact:
Re: Elmer-8.3 -- Maximum Output Level 40 gives huge output
Hi
This is unfortunate. We should minimize it.
The maximum level is 32. A typical level should be ~5, and a verbous level ~10. 32 should be used only when you want all possible low level messages.
-Peter
This is unfortunate. We should minimize it.
The maximum level is 32. A typical level should be ~5, and a verbous level ~10. 32 should be used only when you want all possible low level messages.
-Peter
-
- Posts: 54
- Joined: 17 Jun 2015, 10:04
- Antispam: Yes
Re: Elmer-8.3 -- Maximum Output Level 40 gives huge output
Hi Peter,
Thanks for your reply! I thought that the Max output level is 42. Anyway in my Elmer version, this happens from Max Output Level 22 onward. As far as I can see, this happens between the two function calls:
CALL StartAdvanceOutput('....Solver','Assembly' ) and CALL Info( '....Solver', 'Assembly done', Level=4 ).
There are many function calls and a DO loop between these two function calls : (DO t=1,Solver % NumberOfActiveElements) - so some function in between is throwing the string "GetNOFActive: Number of active elements: xxxx" to std output when the Max output level is 22 or more.
-kumar
Thanks for your reply! I thought that the Max output level is 42. Anyway in my Elmer version, this happens from Max Output Level 22 onward. As far as I can see, this happens between the two function calls:
CALL StartAdvanceOutput('....Solver','Assembly' ) and CALL Info( '....Solver', 'Assembly done', Level=4 ).
There are many function calls and a DO loop between these two function calls : (DO t=1,Solver % NumberOfActiveElements) - so some function in between is throwing the string "GetNOFActive: Number of active elements: xxxx" to std output when the Max output level is 22 or more.
-kumar
-
- Site Admin
- Posts: 4826
- Joined: 22 Aug 2009, 11:57
- Antispam: Yes
- Location: Espoo, Finland
- Contact:
Re: Elmer-8.3 -- Maximum Output Level 40 gives huge output
Hi
I lowered the GetNOFActive verbosity level to 32. I.e. the lowest possible.
Basically it would be smartest to have code like
NoActive = GetNOFActive()
DO t=1,NoActive
....
! within the loop only use NoActive, and not the function call.
END
-Peter
I lowered the GetNOFActive verbosity level to 32. I.e. the lowest possible.
Basically it would be smartest to have code like
NoActive = GetNOFActive()
DO t=1,NoActive
....
! within the loop only use NoActive, and not the function call.
END
-Peter
Re: Elmer-8.3 -- Maximum Output Level 40 gives huge output
I am having similar issues, but the "GetNOFActive: Number of active elements: xxxx" statements show up regardless of what I choose for Max Output Level. I generally use Max Output Level = 5, but they are showing up for lower values as well.
I am using the Navier Stokes and Advection Diffusion solvers, but the issues seem to only happen within the Navier Stokes solver. I send the output to a log file, and with my combination of mesh density and number of iterations, I'm getting gigabytes of log files with Navier Stokes turned on (executed Always) and well under 1 MB with Navier Stokes turned off (executed Never). And all the writing to file can't be good for efficiency.
I'm not attaching my log files (for obvious reasons) but they look pretty similar to the one posted earlier in this thread.
I am running version 8.3, installed from the PPA repository (8.2-201707031146~106b9dd~ubuntu16.04.1) and I hadn't noticed this issue in previous releases (5 June 2017 or earlier).
Thanks,
Joe
I am using the Navier Stokes and Advection Diffusion solvers, but the issues seem to only happen within the Navier Stokes solver. I send the output to a log file, and with my combination of mesh density and number of iterations, I'm getting gigabytes of log files with Navier Stokes turned on (executed Always) and well under 1 MB with Navier Stokes turned off (executed Never). And all the writing to file can't be good for efficiency.
I'm not attaching my log files (for obvious reasons) but they look pretty similar to the one posted earlier in this thread.
I am running version 8.3, installed from the PPA repository (8.2-201707031146~106b9dd~ubuntu16.04.1) and I hadn't noticed this issue in previous releases (5 June 2017 or earlier).
Thanks,
Joe
-
- Site Admin
- Posts: 4826
- Joined: 22 Aug 2009, 11:57
- Antispam: Yes
- Location: Espoo, Finland
- Contact:
Re: Elmer-8.3 -- Maximum Output Level 40 gives huge output
Hi Joe,
I wonder why the output level does not work. Anaways, I removed all the calls of style AdvanceOutput(n,GetNOFActive) that probably were the root cause of the problems. Available in fresh devel branch.
-Peter
I wonder why the output level does not work. Anaways, I removed all the calls of style AdvanceOutput(n,GetNOFActive) that probably were the root cause of the problems. Available in fresh devel branch.
-Peter