Manual and FAQ
1. How to request account?
The account request process works like this:

  • Request an account by filling out and submitting the form.
  • Click the icons for help on any item. The red asterix * denotes a required item.
  • Upon submission you will receive an automated confirmation email with your request number from

2. How to login to the a2c2 system?
Logging in to saguaro can be done through various methods: without GUI(X11) forwarding
Steps for different operating systems is available in HOWTO->Get Started -> Connect to the supercomputer.
Logging into Saguaro with X forwarding and with X11 forwarding via NX Machine can be viewed under HOWTO-> Getting Started.

3. Forgot user password?
Send a password change request to

4. How many CPU hours of my project have been used?

["saguaro1:~]$ mybalance
Project Machines Balance 
———- ——– ——— 
HPCI ANY 394740.71
cse598s112 ANY 4700.28

5. How to add cpu hours?
CPU hours can be purchased by filling the form under and request.

6. How to copy files to/from the local system?
File transfer can be done through scp on Linux
for downloading:

scp user@remote_host:remote_file local_file 

for uploading:

scp local_file user@remote_host:remote_file 

7. How to find my scratch directory?
Scratch space has to be requested in order to access it. Send in the mail to
Find the scratch space with:

cd /scratch/username

1. How to submit, add or delete a batch job?
Batch job can be submitted using PBS script with various parameters. Find the example batch scripts and detailed explanation under Running jobs->Submitting a batch job:
After the script submit the job using:
qsub script.pbs 

Deletion of jobs can be done:

qdel job_id 

2. How to check the status of the system or free compute nodes?
Status of the system can be checked using:


3. Maximum time of the job?
Maximum wall time for interactive job is 24 hours.
Wall time can be mentioned in the PBS script and differs for various queues

4. Combination of nodes and ppn to a run a multi-threaded application?


this means no. of nodes is 2 and no. of processors per node is 4.Thread does not have affinity with nodes. Generally each thread is assigned each core. More than one can also be allocated but one-to-one is the best practice.

5. Job fail after running for few hours?
Jobs may fail due to various reasons: exceeding the walltime limit, wrong combination of nodes and processors.

6. How to increase the walltime of a running job?

qalter -lwalltime=new_walltime jobid 

7. Restoration of deleted files?
Files once deleted cannot be restored. Scratch space is purged once every 30 days.

8. How to get an email notification for any related to submitted job?

#PBS -m abe 
#PBS -M 

PBS can send informative email messages to you about the status of your job. Specify a string which consists of either the single character “n” (no mail), or one or more of the characters “a” (send mail when job is aborted), “b” (send mail when job begins), and “e” (send mail when job terminates). The default is “a” if not specified. you should also specify the email address to which the message should be send via the -M option.

9. How to see the list of specific modules and load them?

$ module load openmpi/1.4.3-intel 
$ module list 
Currently Loaded Modulefiles: 
1) intel/12.0 
2) openmpi/1.4.3-intel(default) 

10. Can we check the memory usage of the job?
Login into the node the current job is running and top command.

11. Can the start time of the job be estimated?
Start time of a job depends on the availability of resources. PBS scheduler schedules the jobs.