First, review this FAQ document to see if your problem is one already answered here.
Second, please refer to the man page for the command(s) with which you experience problems. Incorrect commands and/or typos are common problems. In order to read a manual page online, use the "man" command followed by the name of the command in which you are interested (e.g. "man lpr" lists the manual page for the command lpr). In addition to the online man pages, the AIX operating systems provide online help books, which are made available via the "info" command. You will need an X server available (either on the console, another UNIX box with X, or a PC/Mac with X emulation software such as eXceed or MacX). IBM plans to make this facility available via HTML in the next version of the OS (removing the need for an X server). Full Sun documentation is available via the Sun Product Documentation Web site.
Finally, if you can't find an answer in this document or if immediate assistance is needed (for example, a system is unavailable), please send e-mail to ptpadmin@psu.edu.
When sending e-mail to ptpadmin@psu.edu, please include the following:
Please use the Subject line of your e-mail message to describe your problem so that it can be assigned to the appropriate person on the ptpadmin mailing list.
Typically, the turnaround time for an answer is within 24 hours. Again, if it is an emergency, it will be handled as soon as possible. If an answer is not provided within 48 hours, please resend your message. The system administrators on ptpadmin@psu.edu make every effort to answer inquiries and problem reports within a 24 hour time frame; however, other priorities and/or the time it takes to research or reproduce problems may cause a longer turnaround time. Sending e-mail to ptpadmin@psu.edu rather than directly to one of the system administrators, copies it to both the systems and applications staff, providing for a faster response rate.
The ITS UNIX Cluster Web page describes the resources/machines.
NOTE: If you had a CAC PTR Account, it was automatically converted to an ITS UNIX Cluster Account (see below for additional details).
To access the cluster machines or armstrong, use secure shell protocol (ssh) version 2 client.
The ITS UNIX Cluster requires the addition of UNIX service to an active Penn State Access Account. The same userid and password pair used for one's Access Account also are used to access the ITS UNIX Cluster machines as the same authentication service is used.
An instructor may obtain the service for his/her students from a class list for instructional purposes. The service also is .available to researchers and faculty members and their students.
Information is found via the ITS Accounts Services Office Web site and an application for the service may be submitted via the online form.
If you have questions or need assistance with the application, contact the ITS Accounts Services Office via e-mail at accounts@psu.edu or by phone at 814-865-4772.
The machine speed depends on the type of job you need to run, the features of the machine and the amount of work, in addition to yours, the machine needs to do. The ITS UNIX Cluster machines are multi-user systems and may run processes from a number of other people while you use them, too.
Please review Why is this machine so slow? What's going on? for more information.
This is called running a job in "background" or "backgrounding a job." There are several ways to do this. The process may be started (called job in the example below) in the background by typing:
The first part of the line is the prompt.
For more information on "nice" see the man page or the next entry on "nice" and job priorities. There also is a "renice" command in the event that you forget to start your job with the "nice" command.
In this example, "job" is the name of the program.
The output of the job, which would normally go to the terminal, is redirected to a file (with the "> outfile" part of the command line).
The ampersand (&) at the end places the job in background and allows the user to have control over the terminal. In some cases, the person who uses a terminal after you will see all the output from your job if you don't redirect it to a file. This is a matter of courtesy, with the side effect that it increases a job's performance if the job writes a lot of output.
You also may background a job you already have started if you use csh or tcsh as your command shell. Simply type a control-z (hold down the control key, type z) on the shell window where the job runs. This "stops" the jobs and allows you to type "bg" or "fg" to restart the job in the background or foreground respectively. This also gives you the process id of the job, which you can use to monitor or kill a job. See the "Job Control" entry of the csh man page for more details on this. Jobs backgrounded in this fashion may be killed when you log out.
The "nice" command allows a user to lower the priority of a job. The Scheduling priority is a number; the higher the value, the lower the command's scheduling priority; the lower the value, the higher the command's scheduling priority. In addition to the nice value, the kernel also uses recent cpu usage by the process, the time the process has been waiting to run and other factors to control scheduling.
A positive nice value lowers the priority of your job so that other jobs may access the cpu at a higher priority. The command is called "nice" because you are being nice to others by invoking your job with the "nice" command. To invoke "nice" for your job "jobname" and redirect the output to file "outfile", type:
It is recommended that you "nice" jobs that will take more that a few minutes to run. If you plan to log out and leave the job running, please redirect the output and possibly the error messages to an output file. If the "nice" command is invoked without a nice value, the default value is used. Remember that the ITS UNIX Cluster machines are used by many, so it pays to be "nice."
There are many factors that can affect a machine's performance: cpu utilization, memory consumption and/or disk I/O and network traffic. To find out why the machine that your are on is running slowly, first check the process table. If the machine you're on runs slowly, use the "ps" command. You can use the "ps" command to get information about the "processes" running on the system. For example:
The commands "lpr" and "lp" are used to print text or postscript files. The command "enscript" is used to print text files. The manual pages (man lp) for the commands provide details.
You can determine the system default printer with the command: "lpstat -d".
You can determine names of the printers known to the system with the command: "lpstat -s".
You can override the system default printer either when you print or by setting the "PRINTER" environment variable.
Charges for printing are handled the same way in which charges for printing in the ITS Student Computing Labs are handled. More information about printing in the ITS labs is found via the Classroom and Lab Computing - Printing Services Web page.
The ITS UNIX Cluster accepts ssh connections from any machine on a Penn State network, the ITS dial-up modems and external ISPs (dial-up modem, cable modem, DSL, etc.). Refer to the names of the machines as noted on the ITS UNIX Cluster page and issue the command "ssh machine" to get a login prompt.
If you send e-mail via an ITS UNIX Cluster workstation, your e-mail messages will automatically include your correct Penn State e-mail return address, xyz123@psu.edu where xyz123 represents your Access Account userid.
Any e-mail messages sent to "xyz123@psu.edu," for example, will forward in accordance with the settings in your directory entry. You can review your current e-mail forwarding address by visiting the ITS Secure Server at https://www.work.psu.edu/ and selecting "Change your e-mail forwarding address" link under Directory Services. The page will display the current setting and allow you to change it. Likewise, your e-mail forwarding address is noted in your Penn State Online Directory entry.
To read your e-mail from these machines, you can use the program popclient, execute Pine and then proceed to check your e-mail. This program moves your e-mail from Penn State's POP server to the access_mail directory in your PASS. If you wish to keep your e-mail, do not remove this directory.
The alternative is to configure your e-mail client to do POP mail from one of the University POP mail servers managed by ITS (email.psu.edu or mail.psu.edu). More information about Penn State's e-mail servers is found in the Electronic Mail section of the Guide to Information Technology Services at Penn State.
The UNIX operating system uses the notion of "PATH". This represents the directories in which your shell looks for a command. If you can't find a command, it is because it isn't in your PATH. To check your path type:
If the command isn't in your path, there are several things you can do to find it. Many of the commands that aren't found in traditional paths have the full path name to the executable in the man page entry for those commands. Some UNIX implementations have the full path to every command in their man pages.
Another useful UNIX utility is the "whereis" command. The "whereis" command locates the binary, source and manual page for a given file name. The man page for "whereis" lists what directories are searched.
At times, local filesystems may fill up. This also can happen if a user sends a particularly large job to a printer or a large mail file that cannot be delivered. If the machine tells you that /var is full ("df /var" indicates that the %utilization is >= 100%), then many of the commands that use /var/tmp as a scratch/tmp space will not work. The easiest thing to do is send e-mail to ptpadmin@psu.edu and note the machine on which you are experiencing the problem and switch to one of the other machines on the Cluster. The /var filesystem is a local filesystem (not shared) on each of the machines on the ITS UNIX Cluster. It is likely that /var won't be full on any of the other machines.
If you exceed your quota, you will need to clear out old files in order to free up some space. If you wish to purchase additional space, you may do so by following the instructions provided by the ITS Accounts Services Office Web site.
The Distributed File System (DFS) is a network distributed filesystem, used to provide storage space across a wide range of services. Basically, it allows you to seamlessly access the same files on multiple machines.
This service is commonly known as Penn State Access Account Storage Space (PASS). What this means to you is that all Penn State computer services using PASS see the same files. Examples include the Penn State Portal, Penn State WebMail, the ITS UNIX Cluster and Personal Web Space.
Because PASS is used to serve Penn State Personal Space Web pages, you can maintain your Web site by using an ITS UNIX Cluster Account.
| Penn State | ITS | ASET | UNIX Group | AIT |
The Pennsylvania State University © 2003. All rights reserved. Tuesday, 14-Sep-2004 12:58:44 EDT |