This is an old revision of the document!
This site provides documentation and resources for the COR@L lab. This is a place where COR@L users can find guides on how to access and use the software available to them. All COR@L users have an account provided to them. Thus, in all of these resources, we assume that you already have a COR@L account. If you have any comments or suggestions, such as recommendations for additional resources to include on this site, please contact COR@L Admins.
Every user should first change their initial password received via mail using command “yppasswd”.
We have the following servers in COR@L. Their purposes are explained next to them.
Web server of COR@L website. Hosts personal websites of users, MOPTA conference, INFORMS student chapter and some other sites related to the department.
We use shark to connect COR@L Lab outside of campus. Only two servers in COR@L Lab can be reached outside of campus, coral and shark. Working on shark is not recommended, it lacks the basic software available on other servers (compilers, mathematical and optimization software, etc.).
Polyps is COR@L Lab’s cluster composed of 16 nodes (polyp1-polyp15, polyp30). It is our heavy duty cluster. Polyps has Torque installed as a job scheduler. Torque aims to maximize resource utilization and high throughput by scheduling jobs submitted by users and allocating resources (CPU, memory) of Polyps to them. For a brief introudction to Torque see Software section. Long lasting experiments should be submitted to polyps through Torque. You can also check the load of Polyp nodes from ganglia web interface.
You can check hardware section on COR@L website for more technical information about servers.
If you are a professor or PhD student in IS&E department and want a website on COR@L server, let your system admins know through e-mail. We can set up a WordPress page for you.
All servers run Debian 8.10 Jessie.
We have C, C++, Fortran compilers available through GCC 4.7.2. GCC also implements shared memory parallel computing with OPENMP 3.1 standard. You can use GCC for your shared memory parallel computations on any polyp node. We also have MPICH2 implementation of MPI distibuted parallel computing standard available on polyps.
Python 2.7 is installed on Polyps.
We have GNU Debugger (GDB) for debugging. Emacs and Vi is available as editors. Emacs comes with C/C++ and Python modes. We also have GNU make installed. Subversion and trac are also available. You can use your home directory and shark for setting Subversion repositories.
We have following optimization and numerical software available on polyps.
Please let your system admins know if any of the software fails. Do not forget to add the failing command or code.
Please let your system admins know if you need software that is not on the list. System admins do not decide the purchase of commercial software, but if it is freely available we will do our best to have it in COR@L Lab.
You can find the list of system admins here. System admins’ duty is to keep servers running and making sure that we have all the software required by users.