Introduction to HPCC Cluster and Linux
16 minute read
HPCC Cluster Overview
The HPCC Cluster (formerly called biocluster) is a shared research computing system available at UCR. The HPCC website is available here.
What Is a Computer Cluster?
A computer cluster is an assembly of CPU units, so called computer nodes that work together to perform many computations in parallel. To achieve this, an internal network (e.g. Infiniband interconnect) connects the nodes to a larger unit, while a head node controls the load and traffic across the entire system.
Usually, users log into the head node to submit their computer requests via
srunto a queuing system provided by resource management and scheduling software, such as SGE, Slurm or TORQUE/MAUI. The queuing system distributes the processes to the computer nodes in a controlled fashion.
Because the head node controls the entire system, users should never run computing jobs on the head node directly!
For code testing purposes, one can log into one of the nodes with
srun --pty bash -land run jobs interactively. Alternatively, one can log into the test node owl via ssh.
- Over 12,000 CPU cores
- 130 Intel, AMD and GPU nodes
- 32-128 CPU cores per node
- 256-4,096 GB of RAM per node
- 64 GPUs including NVIDIA K80, P100, A100 and H100
- HDR IB @200Gbs
- Parallel GPFS storage system with 5.0 PB usable space
- File system scales to over 50 PB
- Backup of same architecture and similar amount
- Computing tasks need to be submitted via
- HPCC Cluster headnode only for login, not for computing tasks!
- Monitor cluster activity:
Log into HPCC Cluster
- Login command on OS X or Linux
ssh -XY email@example.com
- Host name:
- User name: …
- Password: …
- Host name:
Additional login information can be found on the corresponding HPCC manuals:
Important Linux Commands
The following provides a short overview of important shell commands. Much more detailed information can be found on HPCC’s Linux tutorials.
List content of current directory
Print current working directory
pwd pwd -P # returns physical location in case one followed symbolic link
Search in files and directories
Delete files and directories
Move and rename files
Copy files from internet to
SCP: via command-line (Manual)
Advantages of this method include: batch up/downloads and ease of automation.
scp file user@remotehost:/home/user/ # From local to remote scp user@remotehost:/home/user/file . # From remote to local
RSYNC: via command-line (Manual)
Advantages of this method include: same as SCP plus differential update options and viewing of directory content.
Print (view) content of remote directory
Download directory or file(s)
rsync -avzhe ssh user@remotehost:~/somedirectory . # -a: recursive archive mode (thus -r not required), also preserves permissions, time stamps, etc # -v: verbose # -z: compress data during transfer # -h: print messages in human-readable format # -e: specifies transfer protocol; using ssh here provides encryption during transfer # --delete: files that were deleted on source will be deleted also in backup-destination # -n: for testing use this dry-run option, but drop '-e ssh' in this case
Upload directory or file(s)
rsync -avzhe ssh somedirectory user@hostname:~/
Check Integrity of Files
To check the integrity of files (e.g. after downloading or copying them),
one can use their hash (checksum) values created by
md5sum. These hash
values are specific to a file and very small in size. If a hash value for a
data file is identical with the hash value of the downloaded copy, then the
downloaded copy is usually identical with the source file. The following routine
assumes that a file named
myfile1.txt was downloaded along with its
*.md5 created for testing). Next, the checksum values are compared.
md5sum myfile1.txt # generates checksum md5sum myfile1.txt > myfile1.md5 # saves checksum to file
md5sum -c myfile1.md5 # checks checksum value
Compare Differences Among Directories
Differences in files and content of two directories can be identified with
diff command. The following also shows how to exclude certain files
in this comparison, here a file called
diff -r --exclude=".git" dir1/ dir2/
STD IN/OUT/ERR, Redirect & Wildcards
* to specify many files
ls output to file
ls > file
Specify file as input to command
command < myfile
Append output of command to file
command >> myfile
STDOUT of one command to another command
command1 | command2
Turn off progress info
command > /dev/null
Pipe output of
grep pattern file | wc
STDERR to file
grep pattern nonexistingfile 2 > mystderr
Homework Assignment (HW2)
See HW2 page here.
Permissions and ownership
List directories and files
The previous command shows something like this for each file/dir:
meaning of this syntax is as follows:
rwx: read, write and execute permissions, respectively
- first triplet: user permissions (u)
- second triplet: group permissions (g)
- third triplet: world permissions (o)
Example for assigning write and execute permissions to user, group and world
chmod ugo+rx my_file
+causes the permissions selected to be added
-causes them to be removed
=causes them to be the only permissions that the file has.
When performing the same operation on many files with subdirectories then one can
-R for recursive behavior.
chmod -R ugo+rx my_dir
Since directories have to be executable the capital
X option can be useful which
applies only to directories but not to files. The following will assign
drwxr-xr-x to directories
-rw-r--r-- to files and hidden files.
chmod -R ugo-x,u+rwX,go+rX,go-w ./* ./.[!.]*
Syntax for changing user & group ownership
chown <user>:<group> <file or dir>
Symbolic links are short nicknames to files and directories that save typing of their full paths.
ln -s original_filename new_nickname
Software and module system
- Over 2,000 software tools are currently installed on HPCC Cluster
- Custom installs in user accounts via various mechanisms, e.g. environment management systems such as conda
- Most common research databases used in bioinformatics are available
- Support of most common programming languages used in research computing
- A module system is used to facilitate the management of software tools. This includes any number of versions of each software.
- New software install requests can be sent to firstname.lastname@example.org.
- To use software manged under the module system, users need to learn using some basic commands. The most common commands are listed below.
Print available modules
Print available modules starting with R
module avail R
Load default module R
module load R
Unload specific module R
module unload R/4.2.0
Load specific R version
module unload R/4.1.2
List loaded modules
Installs and package management with Conda:
Big data storage
Each user account on HPCC Cluster comes only with 20GB of disk space. Much more disk space is
available in a dedicated
bigdata directory. How much space depends on the subscription
of each user group. The path of
bigdata-shared is as follows:
All lab members share the same bigdata pool. The course number
gen242 is used as
for user accounts adminstered under GEN242 (here /bigdata/gen242/shared).
The disk usage of
bigdata can be monitored on the HPCC Cluster Dashboard.
HPCC Cluster uses
Slurm as queuing and load balancing system. To control user traffic, any
type of compute intensive jobs need to be submitted via
srun (see below) to the computer
nodes. Much more detailed information on this topic can be found on these sites:
- UCR HPCC Manual
- Slurm Documentation
- Torque/Slurm Comparison
- Switching from Torque to Slurm
- Slurm Quick Start Tutorial
Job submission with
Print information about queues/partitions available on a cluster.
Compute jobs are submitted with
sbatch via a submission script (here
The following sample submission script (
script_name.sh) executes an R script named
#!/bin/bash -l #SBATCH --nodes=1 #SBATCH --ntasks=1 #SBATCH --cpus-per-task=1 #SBATCH --mem-per-cpu=1G #SBATCH --time=1-00:15:00 # 1 day and 15 minutes #SBATCH --email@example.com #SBATCH --mail-type=ALL #SBATCH --job-name="some_test" #SBATCH -p batch # Choose queue/parition from: intel, batch, highmem, gpu, short Rscript my_script.R
STDERROR of jobs will be written to files named
slurm-<jobid>.out or to a custom file specified under
#SBATCH --output in
the submission script.
Interactive sessions with
This option logs a user in to a computer node of a specified partition (queue), while Slurm monitors and controls the resource request.
srun --pty bash -l
Interactive session with specific resource requests
srun --x11 --partition=short --mem=2gb --cpus-per-task 4 --ntasks 1 --time 1:00:00 --pty bash -l
--mem limits the amount of RAM,
--cpus the number of CPU
--time the time how long a session will be active. Under
--parition one can choose among different queues and node architectures.
Current options under
--partition for most users of the HPCC cluster are:
short. The latter has a time limit of 2 hours. Note,
--x11 will only work when logged in with X11 support.
This requires the
-X argument when logging in via
ssh (see above). On OS X system X11 support is provided
by XQuartz which needs to be installed and running on a system prior to loging in to
a remote system. If X11 support is not available or broken then one can still connect via
srun by dropping the
--x11 argument form the
Monitoring jobs with
List all jobs in queue
List jobs of a specific user
squeue -u <user>
Print more detailed information about a job
scontrol show job <JOBID> scontrol show jobid -dd <JOBID>
Custom command to summarize and visualize cluster activity
Deleting and altering jobs
Delete a single job
scancel -i <JOBID>
Delete all jobs of a user
scancel -u <username>
Delete all jobs of a certain name
scancel --name <myJobName>
Altering jobs with
scontrol update. The below example changes the walltime (
<NEW_TIME>) of a specific job (
scontrol update jobid=<JOBID> TimeLimit=<NEW_TIME>
Resourse limits for users can be viewed as follows.
sacctmgr show account $GROUP format=Account,User,Partition,GrpCPUs,GrpMem,GrpNodes --ass | grep $USER
Similarly, one can view the limits of the group a user belongs to.
sacctmgr show account $GROUP format=Account,User,Partition,GrpCPUs,GrpMem,GrpNodes,GrpTRES%30 --ass | head -3
The following list includes examples of several widely used code editors.
- Vi/Vim/Neovim: Non-graphical (terminal-based) editor. Vi is guaranteed to be available on any system. Vim and Nvim (Neovim) are the improved versions of vi.
- Emacs: Non-graphical or window-based editor. You still need to know keystroke commands to use it. Installed on all Linux distributions and on most other Unix systems.
- Pico: Simple terminal-based editor available on most versions of Unix. Uses keystroke commands, but they are listed in logical fashion at bottom of screen.
- Nano: A simple terminal-based editor which is default on modern Debian systems.
- Atom: Modern text editor developed by GitHub project.
Why does it matter?
To work efficiently on remote systems like a computer cluster, it is essential
to learn how to work in a pure command-line interface. GUI environments like
RStudio and similar coding environments are not suitable for this. In addition,
there is a lot of value of knowing how to work in an environment that is not
restricted to a specific programming language. Therefore, this class embraces
RStudio where it is useful, but for working on remote systems like HPCC Cluster, it
uses Nvim and Tmux. Both are useful for many programming languages.
Combinded with the
nvim-r plugin they also provide a powerful command-line working
environment for R. The following provides a brief introduction to this environment.
The following opens a file (here
myfile) with nvim (or vim)
nvim myfile.txt # for neovim (or 'vim myfile.txt' for vim)
Once you are in Nvim, there are three main modes: normal, insert and command mode. The most important commands for switching between the three modes are:
ikey brings you from the normal mode to the insert mode. The latter is used for typing.
Esckey brings you from the insert mode back to the normal mode.
:key starts the command mode at the bottom of the screen.
Use the arrow keys to move your cursor in the text. Using
Fn Up/Down key allows to page through
the text quicker. In the following command overview, all commands starting with
: need to be typed in the command mode.
All other commands are typed in the normal mode after pushing the
Important modifier keys to control vim/nvim
:w: save changes to file. If you are in editing mode you have to hit
:q: quit file that has not been changed
:wq: save and quit file
:!q: quit file without saving any changes
Useful resources for learning vim/nvim
Terminal-based Working Environment for R: Nvim-R-Tmux.
Tmux is a terminal multiplexer that allows to split terminal windows and to detach/reattach to
existing terminal sessions. Combinded with the
nvim-r plugin it provides a powerful command-line working
environment for R where users can send code from a script to the R console or command-line.
Both tmux and the
nvim-r plugin need to be installed on a system. On HPCC Cluster both are configured
in each user account. If this is not the case then follow the quick configuration instructions given in the following subsection.
Quick configuration in user accounts of UCR’s HPCC
Skip these steps if Nvim-R-Tmux is already configured in your account. Or follow the detailed instructions to install Nvim-R-Tmux from scratch on your own system.
- Log in to your user account on HPCC and execute
install_nvimRtmux). Alternatively, follow these step-by-step install commands.
- To enable the nvim-R-tmux environment, log out and in again.
- Follow usage instructions of next section.
Basic usage of Nvim-R-Tmux
The official and much more detailed user manual for
Nvim-R is available here.
The following gives a short introduction into the basic usage of Nvim-R-Tmux:
1. Start tmux session (optional)
Note, running Nvim from within a tmux session is optional. Skip this step if tmux functionality is not required (e.g. reattaching to sessions on remote systems).
tmux # starts a new tmux session tmux a # attaches to an existing session
2. Open nvim-connected R session
*.Rmd file with
nvim and intialize a connected R session with
\rf. This command can be remapped to other key combinations, e.g. uncommenting lines 10-12 in
.config/nvim/init.vim will remap it to the
F2 key. Note, the resulting split window among Nvim and R behaves like a split viewport in
vim meaning the usage of
Ctrl-w w followed by
Esc is important for navigation.
nvim myscript.R # or *.Rmd file
3. Send R code from nvim to the R pane
Single lines of code can be sent from nvim to the R console by pressing the space bar. To send
several lines at once, one can select them in nvim’s visual mode and then hit the space bar.
Please note, the default command for sending code lines in the nvim-r-plugin is
\l. This key
binding has been remapped in the provided
.config/nvim/init.vim file to the space bar. Most other key bindings (shortcuts) still start with the
\ as LocalLeader, e.g.
\rh opens the help for a function/object where the curser is located in nvim. More details on this are given below.
Important keybindings for nvim
The main advantages of Neovim compared to Vim are its better performance and its built-in terminal emulator facilitating the communication among Neovim and interactive programming environments such as R. Since the Vim and Neovim environments are managed independently, one can run them in parallel on the same system without interfering with each other. The usage of Neovim is almost identical to Vim.
\rf: opens vim-connected R session. If you do this the first time in your user account, you might be asked to create an
~/. If so approve this action by pressing
spacebar: sends code from vim to R; here remapped in
:vsplit: splits viewport (similar to pane split in tmux)
gz: maximizes size of viewport in normal mode (similar to Tmux’s
Ctrl-a zzoom utility)
Ctrl-w w: jumps cursor to R viewport and back; toggle between insert (
i) and command (
Esc) mode is required for navigation and controlling the environment.
Ctrl-w r: swaps viewports
Ctrl-w =: resizes splits to equal size
:terminal: opens terminal in split mode or in a separate window, respectively.
:resize <+5 or -5>: resizes height by specified value
:vertical resize <+5 or -5>: resizes width by specified value
Ctrl-w K: toggles between horizontal/vertical splits
Ctrl-spacebar: omni completion for R objects/functions when nvim is in insert mode. Note, this has been remapped in
init.vimfrom difficult to type default
:set mouse=n: enables mouse support and
: set mouse-=ndisables it
:set mouse=a: enables mouse support and
: set mouse-=adisables it. The
aoption applies it to all vim modes, which sometimes results in unexpected behavior. For some users the
noption may be more suitable.
:h nvim-R: opens nvim-R’s user manual; navigation works the same as for any Vim/Nvim help document
:Rhelp fct_name: opens help for a function from nvim’s command mode with text completion support
Ctrl-s and Ctrl-x: freezes/unfreezes vim (some systems)
An extendedi listing of important nvim commands is here.
File browser built into vim:
NERDtree provides file browser functionality for Vim. To enable it, the NERDtree plugin needs to be installed. It is included in the account configuration
Install_Nvim-R_Tmux mentioned above. To use NERDtree, open
a file with vim/nvim and then type in normal mode
The basic NERDtree usage is explained here.
Important keybindings for tmux
Ctrl-a %: splits pane vertically
Ctrl-a ": splits pane horizontally
Ctrl-a o: jumps cursor to next pane
Ctrl-a Ctrl-o: swaps panes
Ctrl-a <space bar>: rotates pane arrangement
Ctrl-a Alt <left or right>: resizes to left or right
Ctrl-a Esc <up or down>: resizes to left or right
Ctrl-a : set -g mouse on: turns on mouse support
Ctrl-a n: switches to next tmux window
Ctrl-a Ctrl-a: switches to previous tmux window
Ctrl-a c: creates a new tmux window
Ctrl-a 1: switches to specific tmux window selected by number
Ctrl-a d: detaches from current session
Ctrl-a s: switch between available tmux sesssions
$ tmux new -s <name>: starts new session with a specific name
$ tmux ls: lists available tmux session(s)
$ tmux attach -t <id>: attaches to specific tmux session
$ tmux attach: reattaches to session
$ tmux kill-session -t <id>: kills a specific tmux session
Ctrl-a : kill-session: kills a session from tmux command mode that can be initiated with
Nvim IDEs for other languages
For other languages, such as Bash, Python and Ruby, one can use the
vimcmdline plugin for nvim (or vim). To
install it, one needs to copy from the
vimcmdline resository the directories
syntax and their files to
user accounts of UCR’s HPCC, the above install script
install_nvimRtmux includes the
vimcmdline (since 09-Jun-18).
The usage of
vimcmdline is very similar to
nvim-R. To start a connected terminal session, one
opens with nvim a code file with the extension of a given language (e.g.
*.sh for Bash or
*.py for Python),
while the corresponding interactive interpreter session is initiated
by pressing the key sequence
\s (corresponds to
nvim-R). Subsequently, code lines can be sent
with the space bar. More details are available here.