Summary

• Two additional ways to interact with the nearline/archive HPSS system: Globus and a Virtual File System.
• Compute jobs on Niagara now have access to temporary burst buffer space.
• Announcement of several techtalks at the SciNet User Group Meetings.

Details can be found below and are also available on the SciNet education website courses.scinet.utoronto.ca and the SciNet wiki docs.scinet.utoronto.ca.

System News

• The Globus interface to the nearline/archive HPSS system is now operational. The endpoint name is computecanada#hpss. See https://docs.scinet.utoronto.ca/index.php/Globus and https://docs.scinet.utoronto.ca/index.php/HPSS for more details.
• A Virtual File System (VFS) interface to the nearline/archive HPSS system is now available on a special dedicated node which you access through the scheduler. This interface makes HPSS seem like a regular file system, which is really only suited for small scale data management. See https://docs.scinet.utoronto.ca/index.php/HPSS for more details.
• Compute jobs on Niagara now get access to temporary space on the Burst Buffer (a faster file system than scratch). For every job on Niagara, the scheduler now creates a temporary directory on the burst buffer called $BB_JOB_DIR. The $BB_JOB_DIR directory will be empty when your job starts and its content gets deleted after the job has finished. The directory is shared among the nodes of a job. This can be useful for I/O heavy jobs whose temporary files do not fit in ramdisk. See https://docs.scinet.utoronto.ca/index.php/Burst_Buffer for more details.
• For consistency with the general purpose clusters Cedar and Graham, the environment variable $SLURM_TMPDIR will be set in Niagara compute jobs. This variable will point to ramdisk, not to local hard drives (since Niagara has none).

Events Coming Up

Unless stated otherwise, all events listed below are free of charge and take place at the SciNet Teaching Room or Boardroom at our offices on the eleventh floor of the MaRS West Tower, suite 1140A (661 University Avenue, Toronto ON M5G 1M1).

Most events are recorded and some are broadcast, but only some of the courses can be taken remotely for SciNet certificate credits.

Registration for SciNet courses should be done by logging into https://courses.scinet.utoronto.ca with your Compute Canada account (the same one that you use to log into Niagara).

• INTRO TO SCINET AND NIAGARA Wednesday February 13, 10:00 am – 11:30 am SciNet Boardroom (suite 1140, 661 University Avenue, Toronto ON M5G 1M1).

  This is a class of approximately 60-90 minutes to introduce SciNet and the new supercomputer Niagara and teach you how to use Niagara.

  Participation counts towards the SciNet HPC Certificate.

  For more information and registration, go to https://courses.scinet.utoronto.ca/419

• SCINET USER GROUP MEETING Wednesday February 13, 12:00 noon – 1:00 pm SciNet Boardroom (suite 1140, 661 University Avenue, Toronto ON M5G 1M1).

  Pizza, user discussion, and a tech talk by Mubdi Rahman, Ph.D, on

  “The Grammar and Tools of Data Visualization in the Era of Big Data”

  For sign up and more information, see https://courses.scinet.utoronto.ca/424

• ADVANCED LINUX SHELL PROGRAMMING Wednesday February 13, 1:00 pm – 4:00 pm SciNet Teaching Room

  Learn how to write bash script, use environment variables, how to control process, and much more. Requires some linux basic command line experience.

  Participation counts towards the SciNet Scientific Computing Certificate.

  For sign up and more information, see https://courses.scinet.utoronto.ca/431

• INTRO TO SCINET AND NIAGARA Wednesday March 13, 10:00 am – 11:30 am SciNet Boardroom (suite 1140, 661 University Avenue, Toronto ON M5G 1M1).

  This is a class of approximately 60-90 minutes to introduce SciNet and the new supercomputer Niagara and teach you how to use Niagara.

  Participation counts towards the SciNet HPC Certificate.

  For more information and registration, go to https://courses.scinet.utoronto.ca/420

• SCINET USER GROUP MEETING Wednesday March 13, 12:00 noon – 1:00 pm SciNet Boardroom (suite 1140, 661 University Avenue, Toronto ON M5G 1M1).

  Pizza, user discussion, and a tech talk TBA.

  For sign up and more information, see https://courses.scinet.utoronto.ca/425

• INTRODUCTION TO THE LINUX SHELL Wednesday March 13, 1:00 pm -4:00 pm SciNet Teaching Room

  Learn the basics of how to use the unix shell in two hours. Very useful for new users of SciNet that have little or no experience with unix or linux.

  Participation counts towards the SciNet Scientific Computing Certificate.

  For sign up and more information, see https://courses.scinet.utoronto.ca/429

• INTRO TO SCINET AND NIAGARA Wednesday April 10, 10:00 am – 11:30 am SciNet Boardroom (suite 1140, 661 University Avenue, Toronto ON M5G 1M1).

  This is a class of approximately 60-90 minutes to introduce SciNet and the new supercomputer Niagara and teach you how to use Niagara.

  Participation counts towards the SciNet HPC Certificate.

  For more information and registration, go to https://courses.scinet.utoronto.ca/420

• SCINET USER GROUP MEETING Wednesday April 10, 12:00 noon – 1:00 pm SciNet Boardroom (suite 1140, 661 University Avenue, Toronto ON M5G 1M1).

  Pizza, user discussion, and a tech talk by Anjali Silva on

  “Bayesian Model-Based Clustering Approaches for Discrete-Valued Gene Expression Data”

  For sign up and more information, see https://courses.scinet.utoronto.ca/425

• ADVANCED LINUX COMMAND LINE Wednesday April 10, 1:00 pm – 4:00 pm SciNet Teaching Room

  Increase you Linux (bash) command line productivity. Requires some basic Linux command line experience.

  Participation counts towards the SciNet Scientific Computing Certificate.

  For sign up and more information, see https://courses.scinet.utoronto.ca/432

• INTRODUCTION TO NEURAL NETWORK PROGRAMMING Starting April 23, 6 weeks, Tuesdays and Thursdays 11:00 am – 12:00 noon SciNet Teaching Room

  This six-week class will introduce neural network programming concepts, theory and techniques. The class material will begin at an introductory level, intended for those with no experience with neural networks, eventually covering intermediate-to-advanced concepts. The programming language will be Python 3.7; experience with Python programming will be assumed. The Keras neural network framework will be used for neural network programming; no experience with Keras will be expected.

  Participation counts towards the SciNet Data Science Certificate.

  For sign up and more information, see https://courses.scinet.utoronto.ca/417

• RELATIONAL DATABASE BASICS Wednesday May 1, 1:00 pm – 4:00 pm SciNet Teaching Room

  Principles and uses of relational databases with practical examples using python and sqlite.

  Participation counts towards the SciNet Data Science Certificate.

  For sign up and more information, see https://courses.scinet.utoronto.ca/433

• INTRO TO SCINET AND NIAGARA Wednesday May 8, 10:00 am – 11:30 am SciNet Boardroom (suite 1140, 661 University Avenue, Toronto ON M5G 1M1).

  This is a class of approximately 60-90 minutes to introduce SciNet and the new supercomputer Niagara and teach you how to use Niagara.

  Participation counts towards the SciNet HPC Certificate.

  For more information and registration, go to https://courses.scinet.utoronto.ca/422

• SCINET USER GROUP MEETING Wednesday May 8, 12:00 noon – 1:00 pm SciNet Boardroom (suite 1140, 661 University Avenue, Toronto ON M5G 1M1).

  Pizza, user discussion, and a tech talk TBA.

  For sign up and more information, see https://courses.scinet.utoronto.ca/427

• INTRODUCTION TO THE LINUX SHELL Wednesday May 8, 1:00 pm -4:00 pm SciNet Teaching Room

  Learn the basics of how to use the unix shell in two hours. Very useful for new users of SciNet that have little or no experience with unix or linux.

  Participation counts towards the SciNet Scientific Computing Certificate.

  For sign up and more information, see https://courses.scinet.utoronto.ca/430

• COMPUTE ONTARIO SUMMER SCHOOL Summer 2019

  The Compute Ontario Summer School on Scientific and High Performance Computing is an annual educational event for graduate/undergraduate students, postdocs and researchers to learn and share knowledge and experience in high performance and technical computing on modern HPC platforms.

  As in previous years, the 2019 Summer School is expected to have three installments, “West”, “Central”, and “East”. Each will be a week-long event with multiple parallel streams.

  More details will be announced later in 2019.

• SCALING TO PETASCALE INSTITUTE August 19-23, 2019

  This will be a free virtual advanced HPC summer school, organized by a number of the US XSEDE sites, also hosted at SciNet. Topics will likely include OpenMP, MPI, CUDA, and OpenACC.

  More details will be announced later in 2019.