CloudLine Metrocluster seminar in the NORD 3 data center: how it was

On September 25, the NORD 3 data center hosted a seminar on the disaster-resistant CloudLine Metrocluster.



The new service was opened by the CEO, Yury Samoilov, who briefly spoke about the main milestones of the company, our network of data centers and, of course, about the cloud services of DataLine.
')


Technical Director Sergey Mishchuk conducted a kind of excursion into the “cloud” history of DataLine, noting the main stages of the CloudLine evolution:

In 2010, the company began to use the Double Take solution for asynchronous file system replication (based on OST and NORD data centers), but this solution was quite expensive.
In 2011, already based on VMware DataLine launches cloud CloudLine. At the same time, taking into account the demand of customers for a disaster-proof solution, tests of the VMware Site Recovery Manager (SRM) system were carried out.
The development of the network infrastructure between the two OST and NORD sites during 2012-2013, as well as the transition to NetApp storage systems, allowed this year to offer an affordable, disaster-proof cloud solution. One of the main differences of the new solution was synchronous replication at the storage level, which makes it possible to keep both sites in an active state (when using the SRM solution, one of the sites remained passive).

Eduard Bavizhev, head of the virtualization team, in his presentation spoke about the architecture of CloudLine Metrocluster and described various failure scenarios:

At the heart of loudLine Metrocluster is the duplication of storage elements within two sites - the data center OST and NORD, the distance is 33 km. The solution runs on NetApp and VMware vSphere storage.
Due to synchronous data replication at the storage level, with failures such as a disruption of the storage links between the data centers or the failure of the storage controller in one of the data centers, the systems will continue to work without interruption. In case of failure of physical servers at one of the sites, the restoration of the virtual environment takes from 2 minutes (this is the time to launch the operating system and virtual machines).
Finally, with the complete failure of one of the data center to restore the system to another site will take up to 15 minutes.

After the question and answer session, the guests of the data center went on a tour of the new NORD 3 data center. In addition to the basic infrastructure systems, visitors could appreciate the design of the data center, which was recently completed.













And from our window the construction of a new data center is visible :)