Many IT managers at Swiss universities are finding themselves confronted with growing amounts of data. As they have to ensure that their universities’ internal IT infrastructure operates both stably and with the necessary flexibility, they are constantly facing new challenges.
What a pain in the neck! The data volume calculated a year ago and the actual data volume now are two very different things. A new research project is generating far more data than anticipated and the user data was also underestimated. Additionally, the decision to now keep all data for four weeks (or even for up to six months if it’s important) as opposed to just two weeks for security reasons is also having an impact. And what’s more, the universities would also like data to be rapidly retrievable, no matter how old it is.
How should a solution that overcomes all these issues be set up in a cost-optimised way to ensure that it will also remain flexible and scalable in the future? First of all, the expected data volume would have to be re-estimated, and then the IT infrastructure expanded. Such a solution often goes hand-in-hand with high semi-variable costs, plus it can take between 8 and 12 weeks for the infrastructure to be physically available. An ample reserve is often included in the calculations, eliminating any need for further expansion in the event of additional growth. Consequently, this solution is neither economically viable nor sufficiently flexible to respond to updates.
Storage systems with rotating hard drives aren’t the only solutions used; tape systems are also often deployed in this context. Due to the technology involved, tape restoration takes a long time – especially if only single objects are involved. What’s more, all the tapes have to remain in the tape library to ensure permanent availability, so outsourcing to another location is impossible.
In recent years, ObjectStore has increasingly become used as a replacement for the likes of tapes so that data can be stored in another location and for a longer period of time. This is ensured by means of an interface (e.g. S3). If backup software such as Veeam or Rubrik integrates this interface, SWITCHengines ObjectStore (S3) can be incorporated with minimal configuration effort. ObjectStore is now used as a scalable memory solution that allows affordable storage of all older backups or even a 1:1 copy of the backup data.
What makes ObjectStore so advantageous is the fact that it offers instant data availability without any delays. Whether you’re starting out with a small data volume (in the gigabytes range) or a larger one involving several hundred terabytes is unimportant.
As SWITCHengines is directly connected to SWITCHlan, universities don’t incur any additional costs – so incoming and outgoing data traffic is always free of charge. Data@SWITCH guarantees that the data is always stored in Zurich or Lausanne. As SWITCH is a foundation that’s wholly owned by the universities, they can exert direct influence on how the SWITCHengines service further develops. These governance mechanisms allow SWITCHengines to guarantee a stable service.
Two price models are available. In the Pay as you Go model, users only pay for the ObjectStore that’s really occupied. This solution is ideal for smaller amounts of data. Contract options are available for certain data volumes (> 50 TB) that are assumed will grow. To use one of these options, a university binds itself for a desired time span (of two to five years) with a minimum data volume and receives a fixed, reduced price for the entire term.
Universities thus have a highly scalable solution that is both future-proof and adapts to their needs in an extremely short space of time.
If you’d like to join the growing number of institutions that would like to implement a solution such as this with SWITCH, we’re more than happy to help you by clarifying everything you need to know in a non-binding appointment. Please visit https://www.switch.ch/engines/enterprise for further information