- The Allocation Process
- Access
- Standard Password
- X.509 Certificate
- Managing your Account
- Changing Your Password
- Your shell
- Email
- Managing your Allocation
- Charging Algorithm
- Checking Your Account Balance
Allocations on the TeraGrid systems are available to eligible
principal investigators through a peer-review process. See
the document
Access to the TeraGrid for details.
A default password is sent to you as part of the new user packet you
receive via postal mail after your account is created. It is important
that you keep the new user packet information that contains your default
password in a secure place for your reference. You will have the same
default password on all NCSA systems. If you already are a user on
NCSA HPC systems, use your current password to log in to the NCSA
TeraGrid cluster. If you are a new NCSA user, use the default password to
log in. You should change your password the first time you log in and
periodically thereafter.
In order to properly authenticate to the NCSA TeraGrid cluster, you
need a version of SSH that has the "keyboard-interactive" authentication
mechanism (most newer versions support this). Windows users should set
up a profile for the cluster, and then edit the authentication mechanisms.
Add keyboard-interactive so that it is the first authentication
mechanism.
Login to tg-login.ncsa.teragrid.org using ssh
to connect to the access node:
% ssh tg-login.ncsa.teragrid.org
Note: If you have a different username on the remote machine, you'll
need to specify it on the command line:
% ssh jdoe@tg-login.ncsa.teragrid.org
You can also use X.509 certificates to authenticate
to the TeraGrid. First you'll need to
get an accepted user
certificate and get added to the grid-mapfile.
See
Access to the TeraGrid: Authentication for information.
Then you can use GSI-Enabled OpenSSH (gsissh) to log onto the access node:
% gsissh tg-login.ncsa.teragrid.org
Note: Because you are using your certificate for authentication, you shouldn't
have to enter a password and you don't have to specify your remote username even
if it is different.
Note: When using GSI-Enabled OpenSSH to connect to the NCSA TeraGrid
cluster, if you want to do anything Kerberos-oriented, such as changing
your password or Kerberos ftp, you'll need to run kinit first.
NCSA requires passwords to be changed via a
secure web page.
When your account is first activated your default shell is set to
tcsh, an enhanced version of the Berkeley UNIX C shell (csh).
The tcsh shell is completely compatible with the standard csh, and
all csh commands and scripts work unedited with tcsh.
Enter "man tcsh" at the prompt for details about
tcsh.
The other shell available on the NCSA TeraGrid cluster is the GNU
Bourne-Again SHell (bash), which is completely compatible with the
Bourne shell (sh).
To change your shell, send email to
TeraGrid Helpdesk with a request.
Email services are not supported on the NCSA TeraGrid system. Outbound
email will be delivered, but replies to these emails will bounce.
In cases where NCSA TeraGrid staff need to contact you, they will send email
to your NCSA account. So it is important that you forward your NCSA
email.
Use the same login and password as when connecting to the NCSA TeraGrid
cluster. At this time, grid services are not running on these machines,
so you will need to know your login and password.
To forward your NCSA email, log onto one of the public email
machines:
- public-linux.ncsa.uiuc.edu
- public-sun.ncsa.uiuc.edu
Use the same login and password as when connecting to the NCSA TeraGrid
cluster. At this time, grid services are not running on these machines,
so you will need to know your login and password.
- create a file /var/forward/$USER/.forward with your favorite
editor which contains a single line with your forwarding address.
- chmod 644 /var/forward/$USER/.forward
Note: Creating a .forward file in your account on the NCSA TeraGrid cluster
does not work. You must create it as described above on the public email
machines.
The Service Units (SUs) charged for a job is based on the number of
nodes used and the wall clock time.
#SUs = 2 * #Nodes * Wall_Time * NormFactor
where
#SUs = the number of service units charged for the job
Wall_Time = Total wall clock hours used by the job
NormFactor = 1 for the 1.3 GHz processor machines
1.15 for the 1.5 GHz processor machines
On the Command Line
For TeraGrid users, you can monitor usage of your allocation by using
the TeraGrid
tgusage
utility.
The data
displayed by tgusage
is normally updated once each day, so SUs accrued by jobs on a given
day will be reflected in tgusage the next day.
Entering tgusage with no options displays cumulative usage
information of all your TeraGrid allocations(projects) on the local
NCSA machine. If you are the Principal Investigator (PI) of the project,
tgusage will return a list of all users in the project along with their
usage. If you are a user in the project, tgusage will return your
specific usage.
Specify the --account option for information about a specific
project.
The usage shown is the individual user's SU usage and total
SU usage for all users in the project.
For a complete description of the tgusage options, enter
tgusage --help.
Online Usage Information
Usage information can also be obtained online through the
TeraGrid User Portal.
- login to the User Portal
- click on the "My TeraGrid" tab
- you should be on the "Accounts and Usage" page (default),
if not click that link under the "My TeraGrid" tab.
NCSA's usage command
Note: Starting Monday February 4 2008, the NCSA "usage" command
will no longer provide TeraGrid allocation and usage
information.
NCSA local and
Private Sector Partner (PSP) users should use usage to monitor usage
of your allocation.
Entering usage with no options displays cumulative information about
the usage of all your allocations on all applicable NCSA machines. For example:
% usage
Proj Mach Login Usage Status P_alloc P_usage P_expire Name
aad abe lahti 91.03 Active 100000 713.72 06/30/08 Lahti, Tracy
non cobalt lahti 0.02 Active 50000 94.93 12/31/08 Lahti, Tracy
If the project name is specified using the -p option,
information is returned
for all users of that project on all applicable machines. The data
displayed by usage
is normally updated once each day, so SUs accrued by jobs on a given
day will be reflected in usage the next day.
For a more complete description of the usage options, enter
man usage.
back to Top