The protection goals of integrity, availability and confidentiality are among the most important and security-relevant factors in cloud computing. Many cloud providers find the availability or high availability of their systems to be a particular challenge. Why? Quite simply: Most of today's technologies have not yet reached the point where they can function without interruption. Even technology leaders, such as Microsoft and Google with their cloud services "Office 365" and "Google Drive", are not prepared for such failures, as the system failures of recent years show.*
German cloud providers in particular are still struggling in this area and therefore prefer to rent IT infrastructure from third parties such as AWS or Microsoft. No wonder, because redundant and scalable data backends in the multi-petabyte range are extremely complex. All the more reason for us to be pleased, of course, that luckycloud GmbH was able to offer an availability of approx. 99.998 % in 2019.
Our own IT infrastructure can be scaled horizontally in parallel with the growth in customer data, practically infinitely. "In 2018, we started to build our own highly available IT infrastructure and that was a very important and, above all, right decision that I made in 2018," Luc explained to me at the time.
luckycloud searched long and hard for a suitable storage backend, and for us the right solution was object-oriented storage, as it is almost infinitely scalable upwards. Storage backends based on Ceph score with performance, reliability and scalability - but these are extremely complex.
One of the biggest challenges in the planning and implementation of the new infrastructure was the many hundreds of millions of small "objects". A lot of small data to trade is not necessarily one of the strengths of distributed storage systems (Ceph). Therefore, some adjustments were necessary to meet our requirements in terms of performance and data redundancy.
Due to the complexity of highly available and distributed storage systems, we were looking for a way to simplify them. This is how we became aware of croit GmbH and finally chose croit. It was really a "perfect match" because croit GmbH, just like luckycloud, is based on open source and we also found a reliable partner who shares the basic values and goals of luckycloud GmbH.
The core motivation of the founding team, Martin Verges, Paul Emmerich and Matthias Grandl, was to provide reliable and easy-to-use management software for Ceph storage clusters.
A brief history: After years of active software development, a lot of work and sleepless nights, a usable prototype was ready by the end of 2016. On 31 January 2017 the company agreements will be signed. The three founders were happy to start together as croit GmbH into the year 2017.
Their mission was to redefine the management of storage clusters. It should be efficient, secure and comfortable. Ceph should be distributed as an open source software defined storage tool.
The stability, performance and availability of luckycloud speak for themselves: croit is not only a suitable cooperation partner for luckycloud because of technical aspects, but there is also always an open and mutual exchange of personal information. Even if the physical distance from Berlin to Munich may seem great, this is not an obstacle in the virtual age today.
Author: Santana Michael
Stefan Beiersmann (2019): Google bestätigt Störung von Gmail und Google Drive. NetMediaEurope Deutschland GmbH
Günter Born (2019): Neuer Ausfall in Microsofts Cloud: Microsoft 365 betroffen. Heise Medien GmbH & Co. KG