User support workshop @ NeIC 2015
This user support workshop is one of several interesting workshops held in connection with the NeIC 2015 conference on May 5-8 2015 in Espoo Finland.
Date and place
Increase knowledge exchange and collaboration among e-Science and HPC infrastructure providers.
- Meet others that work with similar things or have similar responsibilities and work interests.
- Inspire others by presenting and/or discussing how you work with addressing various types of support challenges.
- Get inspired by others by hearing how they approach or tacke various types of support challenges.
- Find interesting projects that you can join or learn from.
If you are working with user-oriented operational services (e.g. software installations, support, etc.), other types of application expertise, training, or other efforts aimed at helping users to utilise e-Science and HPC services, then this workshop is for you.
|Time||Presenter / Activity||Breakout activity|
|9:00 - 9:10||Welcome|
|9:10 - 9:40||Malgorzata Krakowian (EGI)|
|9:40 - 10:20||Neil Chue Hong (SSI)|
|10:20 - 10:30||Networking exercise|
|10:30 - 11:00||Coffee Break|
|11:00 - 11:30||Rossen Apostolov (SNIC/PDC)|
|11:30 - 12:30||Kenneth Hoste (HPC UGent)|
|12:30 - 13:30||Lunch|
|13:30 - 14:30||Christian Kniep (QNIB)|
|14:30 - 14:45||Jonas Lindemann (SNIC/Lunarc)|
|14:45 - 15:00||Åke Sandgren (SNIC/HPC2N)|
|15:00 - 15:30||Coffee Break|
|15:30 - 16:30||Discussions||EasyBuild hands-on|
|16:30 - 17:00||Discussions wrap-up||EasyBuild hands-on|
|17:00 - 18:00||Break||EasyBuild hands-on|
|18:00 - 21:00||Conference opening get-together & barbeque|
- Malgorzata Krakowian
- Supporting the supporters: EGI solutions for federated e-infrastructures and research infrastructures [pptx]
- Neil Chue Hong
- Activities and experiences from the UK Software Sustainability Institute [pptx]
- Rossen Apostolov
- Scientific software development infrastructure [pdf]
- Kenneth Hoste
- Introduction to EasyBuild and Lmod [pdf]
- Christian Kniep
- Docker, monitoring and SLURM Specific visualisations [pdf] [youtube]
- Jonas Lindemann
- Demo and presentation of the Lunarc desktop
- Åke Sandgren
- Software accounting prototype [pdf]
You can either do the hands-on as a breakout activity instead of the discussions part of the main workshop programme or join the hands-on after the main workshop has ended.
The only requirement for the hands-on is that you bring a laptop. However, if you want to shorten the startup a little, you can prepare your laptop by installing VirtualBox, so you are ready to launch the VM image that Kenneth will provide.
If you have access to an HPC system (Linux/x86_64 based, or Cray), then Kenneth will also be able to get you started with EasyBuild in this context. However, please note that this option may be somewhat more time-consuming.
Write focus area tags:
Where focus area is one of the four items described under focus areas below.
Have coffee and find your focus area peers!
What do users need to do, to get access to the infrastructure and what do they need to do for accessing it for actual use?
- Procedure for obtaining a grant (or similar) to use the infrastructure
- Procedure for 1st hands-on access
- Procedure for regular hands-on access
- Ease of use vs. security considerations
What do users need to do, to start producing scientific results? How much of the user’s total workflow is available within the infrastructure?
- Readily available scientific software
- Readily available build environments
- Readily available code development tools (debuggers, profilers, etc.)
- Procedure for getting help
- Type and amount of help available
- Availability of pre- and post-processing tools
- Availability of and ease of use for graphical applications
- Data handling and storage options (backups, file copying, file system mounting, etc.)
- Instant access availability vs. queuing
How efficient are the resources being utilised? How well is the user informed about his/her resource utilisation?
- Monitoring and logging of resource utilisation
- Automatic feedback to users regarding resource utilisation
- Information about OOM events
- Per job resource utilisation reports
- Monitoring and logging of software utilisation
- Rule-of-thumb run-time settings for widely used software
How efficient is the scientific code that the user is running?
- Using the best programming techniques and libraries
- Code maintainability
- Keeping up with new technology
We will divide into groups of 4-6 persons based on focus area.
Take inspiration from the presentations that you have heard today or from other activities/projects that you know of, for example from your home site. Discuss the following general question and try to solve one (or more) of the proposed tasks below.
Create a poster out of one or several flip chart pages addressing one of the items below:
Try to find activities, projects, etc. that the sites represented in your group could share information and knowledge regarding, or potentially even collaborate on. Also describe how you would try to help each other.
Describe what you would need to change within your research infrastructure(s) to service ten times more users with the same staff.
Describe initiatives/implementations that you would propose to improve accessibility for your research infrastructure(s).
Describe initiatives/implementations that you would propose to improve usability for your research infrastructure(s).
Describe initiatives/implementations that you would propose to maximise compute resource utilisation.
Describe initiatives/implementations that you would propose to help users use and develop more efficient code.
National as well as international HPC and e-Science infrastructures are tasked with providing hardware resources and services to, in many cases, a wide and varied range of scientific domains and research disciplines within domains. To ensure that the provided resources are properly accessible to researchers and well utilised, several types of user support efforts are required.
This workshop will aim at taking a broad view on user support efforts and proposed examples include how to organise support, operational services to enable utilisation, training efforts, project support, proactive efforts, and other efforts that enable the utilisation of the infrastructure.
Proposed topics include (but are not limited to):
- Organisational examples
- Organise per discipline
- First line, second line, etc. style
- Visits, satellite persons
- Use of many-to-many support frameworks (e.g. Stack Overflow style)
- Operational services
- Maintaining development environment
- Maintaining scientific software
- Frameworks for providing documentation and user instructions
- Software usage monitoring and statistics
- Resource usage monitoring and statistics
- Project support
- Code profiling
- Code porting
- Code optimisation
- Proactive efforts
- Automatic OOM alerts
- Per job resource utilisation reports
- New technology early access and help (e.g. accelerators)