Aura

Server aura.fi.muni.cz is available to FI staff and students for longer or more demanding calculations. Access to the machine is created automatically by FI and PhD students. For study or research purposes, FI staff may grant access to other persons through faculty administration . If you are planning a more extensive use of Aura (multiple students) for teaching, please contact first unix@fi.muni.cz .

Way of work

We are not currently considering the introduction of a queue system for long-running tasks. It is enough to run long-running processes with a reduced priority (in the range of 10-19, 19 is the lowest). Long-running processes are those that run for more than an hour. To do this, for example, use the command nice ./váš_program . Additionally, you can restrict the program to use only certain processors with the command taskset . Example of starting with minimum priority on the 11th to 15th processor:

nice -n 19 taskset -c 10-14 ./váš_program

You can use the command to change the priority of an already running process renice which can be applied to threads, processes or groups of processes (see the man page of the command). However, you need to keep in mind that a process can run in multiple threads, and changing the priority for one process can change the priority of only one thread. A list of all threads in my processes, including priority, can be obtained, for example, with the following command:

ps x -Lo pgid,pid,tid,user,nice,tty,time,args

You can perform short-term processes or interactive debugging of your programs with normal priority.

If your process does not comply with the priority limit and uses a lot of computing power, the lowest priority will be set to prevent other users from restricting all your processes 19 . Repeated or more serious violations of this rule may result in the suspension of your faculty account.

For memory-intensive calculations, be careful not to run out of available memory and thus invoke the OOM killer mechanism, which can terminate another user's process. The memory group setting for all user processes in the summary is crucial for the usable memory limit. You can find out its value with the command:

systemctl show -p MemoryLimit user.slice

You may also find the following commands useful for tracking or limiting resource usage:

# monitoring of the memory and CPU usage of your processes
systemd-cgtop /user.slice/user-$UID.slice

# limit available resources
help ulimit
# cap the size of virtual memory to 20000 kB
ulimit -v 20000
# cap the amount of total CPU time to seconds
ulimit -t 3600
# cap the number of concurrently running threads/processes
ulimit -u 100

All of the above commands limit the resources of the shell and all its descendants to the specified values. These cannot be rolled back, you will need to run another separate shell to restore the environment without any restrictions.

Specific software

If you need to install libraries or tools for your work, you have (apart from local compilation) several options:

• if they are part of a distribution ( dnf search software-name ), you may request administrator about installation,
• you can make module ,
• as for the Python package, it can also be installed locally using pip/pip3 install --user , if applicable virtualenv .

Disk capacity

Directories are available for temporary data that should be available locally quickly /tmp and /var/tmp on a fast RAID 10 volume. To use this space, store your data in the directory with your login. Unaccessible data ( atime ) are automatically lubricated, u /tmp at the age of several days, u /var/tmp in months (the exact settings can be found in the file /etc/tmpfiles.d/tmp.conf ). Disk quotas do not apply here; however, be considerate of others in your use of space.

Hardware configuration

Server Aura now runs on HP DL980 G7 platform in configuration:

• 8 eight-core 64-bit Intel Xeon X7560 2.26 GHz processors ( 64 cores in total).
• 448 GiB DDR3 RAM
• Gigabit Ethernet connection
• 8 disks with a capacity of 300 GB and 12 disks with a capacity of 1.8 TB, connected in RAID 10, ie 12 TB of logical capacity