Welcome, researchers!

Welcome to Aalto, researchers. Aalto has excellent resources for you, but it can be quite hard to learn of them. These pages will provide a good overview of IT services for researchers for you (focused on computation and data-using researchers).

These aren’t generic IT instructions - ITS has an introduction for staff somewhere (but apparently not online). There is also a handy list of research-focused services provided by ITS.

Note

Under construction

Aalto service units

Understanding all the Aalto services can be quite confusing. Here are some of the key players:

  • Department IT: Only a few departments (mainly in SCI) have their own IT staff. Others have people such as laboratory managers which may be able to provide some useful advice. Known links: CS, NBE, PHYS.
  • Science-IT: Overlaps with SCI department IT groups. They run the Triton cluster and support scientific computing. The main target audience is the departments which fund them, but can assist anyone in the university. The core Science-IT departments are CS, NBE, and PHYS. Science-IT runs a weekly SciComp garage, where we provide hands on support for anything related to scientific computing. Official site, but this site is its day-to-day guidance.
  • Aalto ITS (IT Services): Provides central IT infrastructure. Their support for complicated research situations and special scientific software is limited, but they are the first place to contact if not in the School of Science. Their infrastructure is also used in SCI, but supported by department teams. Their instructions are on Aalto Inside, but most importantly the already-mentioned IT Services for Research page. Contact via servicedesk.
  • Research and Innovation Services: Administrative-type support. Provides support for grantwriting, innovation and commercialization, sponsored projects, legal services for research, and research infrastructures.
  • CSC is the Finnish academic computing center. They provide a lot of basic infrastructure you use without knowing it, as well as computing and data services to researchers (all for free).

Also, currently Aalto has information is scattered everywhere:

  • aalto.fi is the normal homepage, but doesn’t have much practical information for researchers.
  • it.aalto.fi is a new location for IT instructions, and somewhat replaces the old IT instructions on inside.aalto.fi and into.aalto.fi. It is still under active development. These generally relate to how to do a specific task - but not necessarily what is the best thing to do.
  • inside.aalto.fi is the typical “official” staff documentation area, where most service units have their homepages. This is perhaps the most important place to look, but is not visible in any public search engines or on the search engine on the Aalto homepage - you have to go directly here to search (so most people never find the Aalto official information).
  • into.aalto.fi is the student official information. It contains a lot of duplicate information compared to Inside and is public, so if you search things at Aalto you tend to find answers here.
  • wiki.aalto.fi is obviously the Aalto wiki space. Anyone can make a space here, and many department’s internal sites are here. Searching can randomly find useful information. Most sites aren’t publically searchable.
  • scicomp.aalto.fi is where you are now. Scicomp was started by the Science-IT group from the Triton (HPC cluster) documentation, which is slowly taking over from our department’s research IT instructions. Now is is our general guidance to researchers. We would rather combine with other official sites, but currently there’s nothing standard-based and open.

End-user systems

Aalto provides computers to it’s employees, obviously. You can choose a managed system or standalone. If it’s standalone, you are on your own. If managed, login is through your Aalto account. You can get laptop or desktop, and Linux, Mac, or Windows.

Desktops are connected directly to the wired networks. Linux desktops have fast and automatic access to all of the university data storage systems, including Triton and department storage. They also have a wide variety of scientific software already ready (and somewhat similar to Triton).

Managed laptops are usable in and out of the Aalto networks, and you can become “primary user” which allows you to install some software yourself (Linux at least).

Computing

You have two primary Aalto options: workstations and Triton. The Aalto workstations have basic scientific software installed. From the workstations, you can use the HTCondor distributed computing framework.

Most HPC at Aalto is performed on Triton, the Aalto HPC cluster. It is a fairly standard medium-sized cluster, and it’s main advantage is the close integration into the Aalto environment: it shares Aalto accounts, its data storage (2PB) is also available on workstations, and has responsive local support. If you need dedicated resources, you can purchase them and they can be managed by us as part of Triton. Triton is part of the Finnish Grid and Cloud Infrastructure.

CSC (the Finnish IT Center for Science) is a government-owned organization. They provide a lot of services, most notably HPC, data, and IT infrastructure services to the academic sector. All of their services are free to the academic community (paid directly by the state of Finland). They also coordinate the Finnish Grid and Cloud Infrastructure. They have the largest known cluster in Finland.

Data

Data management isn’t just storage: if data is just put somewhere, you get a massive mess and data isn’t usable in even 5 years. Funders now require “data management plans”. Thus data management is not just a hot topic, it’s an important one. We have a whole section on data, and also there are higher level guides from Aalto, or you can check our guide for researchers. If you have specific questions, there is an official service email address you can use (see the Aalto pages), or you can ask the Science-IT team.

Aalto has many data storage options, most free. In general, you should put your data in some centralized location shared with your group: if you keep it only on your own systems, the data dies when you leave. We manage data by projects: a group of people with shared access and a manager. Groups provide flexibility, sharing, and long-term management (so that you don’t lose or forget about data every time someone leaves). You can request as many projects as you need, and each can have its own access control and quota. You can read about the storage locations available and storage service policy.

Triton has 2PB of non-backed up data storage on the high-performance Lustre filesystem. This is used for large active computation purposes. The Triton nodes have an incredible bandwidth to this and it is very fast and parallel. This is mounted by default at Science-IT departments, and can be by default in other departments too.

Aalto provides “work” and “teamwork” centralized filesystems which are large, backed up, snapshotted, shared: everything you may want. Within the Science-IT departments, Science-IT and department IT manages it and provides access. For other schools/departments, both are provided by Aalto ITS but you will have to figure out your school’s policies yourself. It’s even possible to directly collect data to these systems from lab equipment.

CSC provides both high-performance Lustre filesystems (like Triton) and archive systems.

In our data management section, we provide many more links to long-term data repositories, archival, and so on. The OpenScience.fi project is state-supported and has a lot more information on data. They also provide some data storage focused on safety and longer-term storage, though they are harder to use.

Software

Triton and Aalto Linux workstations come with a lot of scientific software installed, with in the Lmod system. If you are the primary user of a workstation, you can install Ubuntu packages yourself (and if you aren’t, you should ask to be). If you use Triton or are in a Science-IT department, it can be worth asking Science-IT about software you need - we are experts in this and working to simplify the mess that scientific software is. Windows workstations can have things automatically installed, but you’ll have to find the method from the ITS instructions above.

Triton and Aalto workstations have the central software available, currently for laptops you are on your own.

On Triton, type module spider $name to search for available software. We are working to unify the software stack available on Triton and Aalto workstations.

ITS has a software and licenses (FI) page, and also a full list of licenses. There is also https://download.aalto.fi/.

CSC also has a lot of software. Some is on CSC computers, some is exported to Triton.

Starting a project

Each time you start a project, it’s worth putting a few minutes into planning so that you create a good base (and don’t end up with chaos in a few years).

  • Do the normal Aalto bureaucratic work. We don’t know about that, and this guide does not relate to that. You also don’t have to do that stuff in order to do the steps below.
  • Request a data group (see above) if you don’t already have a good location. This will keep all of your data together, in the same place.
    • If you already have a data group that is suitable (similar members), you can use that. But there’s no limit to the number of projects, so think about if it’s better to keep things apart earlier.
    • Mail your department IT support and request a group. Give the info requested at the bottom of data outline page.
    • In the same message, request the different data storage locations, e.g. scratch, project, archive. Quotas can always be increased later.
  • Think about how you’ll manage data. It’s always easy to just start working, but it can be worth getting all project members on the same page about where data will be stored and what you want to happen to it in the end. Having a very short thing written will also help a lot to get newcomers started. The “practical DMP” section here can help a lot - try filling out that A4 page to consider the big sections.

Training

Of course you want to get straight to research. However, we come from a wide range of backgrounds and we’ve noticed that missing basic skills (computer as a tool) can be a research bottleneck. We have constructed a multi-level training plan so that you can find the right courses for your needs. These courses are selected by researchers for researchers, so we make sure that everything is relevant to you.

Check our upcoming training page for a list of upcoming courses. If you do anything computational or code-based at all, you should consider the twice-yearly CodeRefinery workshops (announced on our page). If you have a Triton account or do high-performance computing or intensive computing or data-related tasks, you want to come to the Summer (3 days) or Winter (1 day) kickstart, which teaches you the basics of Triton and HPC usage (we say it is “required” if you have a Triton account).

Other notes

Remember to keep the IT Services for Research page close close at hand!

Research is usually collaborative, but sometimes you can feel isolated - either because you are lost in a crowd, or far away from your colleagues. Academic courses don’t teach you everything you need to be good at scientific computing - put some effort into working together with, learning from, and teaching your colleagues and you will get much further.

There are some good cheatsheets which our team maintains. They are somewhat sepcialized, but useful in the right places.

It can be hard to find your way around Aalto, the official campus maps and directions are known for being confusing confusing. Try UsefulAaltoMap instead.