To run Accelrys' InsightII Release 97.0/4.0.0 on Athena:
- Log in to any Athena Silicon Graphics Indy workstation.
- Type
add biosym_root to connect the Biosym locker (InsightII software was originally developed by Biosym Technologies).
-
- For Materials Science applications, type
source /mit/biosym_root/cshrc400 to set up the appropriate environment variables, etc.
- For Life Science applcations, type
source /mit/biosym_root/cshrc970 to set up the appropriate environment variables, etc.
- Type
insightII to fire up the InsightII user interface.
There is on-line documentation available for the Accelrys software (for MIT use only). To get the ID and password to access this documentation, you must first add the biosym_root locker, and then type more /mit/biosym_root/README.web-password. (In this module we are using the InsightII 97.0 release and the Polymer 400P release.)
Release Notes:
- There are online tutorials which you can prepare by running "biosym_tutorial" after the appropriate "source" command. You can then select Helpt -> Online Tutorials from the insightII interface. If you run tutorials you will have a lot of small files in a "tutorial" subdirectory of your home directory- these are mostly links into the Biosym tree and can be deleted anytime.
- If you have been running one of the releases, such as 400, and want to switch to the other (i.e. 970), simply type "set_biosym_context 970" at the UNIX prompt. It will change a bunch of paths and environment variables which will get you to 400. You can do the reverse to go the other way.
Troubleshooting
During startup of insightII, you may get a message like the following:
>>dks0d1s0: Process [insightII] ran out of disk space [HELP=DISK_FULL]
>>dks0d1s0: Process [insightII] ran out of disk space [HELP=DISK_FULL]
>>run: unable to write temporary file: No space left on device
This is because the temporary space InsightII needs on the local disk
while it runs has filled up. In fact this could happen with any big
application, not just InsightII. This was an issue on the SGI Indys when they
were in service. Workstations which run an automatic flush of the temp areas
should not experience this problem.
A work-around for now in the event that this happens is to manually delete:
- All files bigger than 5K owned by them in /usr/tmp and /tmp
- All files whose names start with unix and vm in /var/adm/crash
In order to to do #2. you will need to su to root. You can find the root
password by typing "tellme root"; it's generally drroot these days.
If you are running long jobs, it may in some cases be wise to do this
before you start a run. It generally takes a day or two for this space to
fill up.