|Posted on Sunday, February 26, 2012 - 10:13 pm: |
We are having an issue where the vibrational data for a molecule will simply not show up. The table where the data normally shows up is just not there.
When trying to view molecular orbitals, the table shows up but the first orbital is populated with an out of bounds array reference and the rest show NULL.
We are using Gaussian 09 for all calculations.
This issue is only occurring when using WebMO 10.1.002 enterprise.
I tried installing WebMO 12.0.003 free version and the problem is no longer present.
With the free version I can view the vibrational data with no problem.
However, the free version does not allow for molecular orbital viewing.
I tried this using Suexec and cgiwrap, but got the same issues either way.
the only clue I could find is in the errors log.
The free version provides no errors.
the enterprise version shows the same error for various cgi scripts:
Array @atoms missing the @ in argument 1 of push() at parse_gamess.cgi line 525.
I am unsure why parse_gamess.cgi is being called at all, since we are using G09 and have not added GAMESS in the interface manager.
I am at a loss here..
Post Number: 230
|Posted on Monday, February 27, 2012 - 10:13 am: |
The error you are seeing in the error log is unrelated, and unimportant.
Do these problems occur with all vibrational calculations, or just this particular one?
Does the MO problem occur with ALL MO calculations, or just this particular one?
Is this a new problem, or has this always occured?
Please send me an example output file that exhibits this problem.
|Posted on Tuesday, February 28, 2012 - 1:30 pm: |
I did some further research and found the following:
On our main server, all vibrational and MO calculations have the problem.
I tried this on our old server and had no problems using suexec and without suexec.
I have reinstalled WebMO several times on each machine and recompiled and installed the apache web server several times as well.
The differences between the two machines other than hardware differences is that the old server is running Slackware 13.1 while the new server is running Slackware 13.37.
I am going to examine the change log for the different version of Slackware to see if that points me in the right direction.
If nothing can be found, I will try installing the new version of Slackware on the old server to see if the problem starts occurring there.
I have attached the output file of a vibrational calculation that shows the error.
I also tried copying an output file that worked fine on the old server to the new server. When it was imported, the vibrational data is again missing.
This indicates that it is not likely to be an issue with the output files that are generated, but rather something in the new server that is not able to read them properly...
Post Number: 231
|Posted on Tuesday, February 28, 2012 - 1:33 pm: |
I know the problem. It is a Perl issue. Check the Perl versions on the two machines, but my guess it that your new server has Perl 5.10 5.12 or newer. Perl 5.10 deprecated some features I had been using, which breaks some things in WebMO. This is resolved in WebMO v12. The changes are numerous, as such you can either:
a) Upgrade to WebMO v12, or
b) Downgrade your perl to 5.8
Either should resolve the issue.
|Posted on Tuesday, February 28, 2012 - 2:20 pm: |
The new server running Slackware 13.37 had Perl 5.12.
The old server had Perl 5.10.1
I downgraded to Perl 5.10.1 and I can now view vibrational and MO data
Thanks for your help