Backup data center

Firstly, business continuity

Despite the economic problems, Poland has been a big construction site in recent years with structural funds from the European Union, preparations for Euro 2012, and a large number of road and infrastructure investments. Each major investment involves the use of hundreds, or even thousands of equipment items, kilometres of fences and scaffoldings, and temporary buildings to house workers. Ramirent participates in the largest construction projects in Poland by providing all types of equipment and machinery – from heavy equipment, telescopic handlers, hoists, containers, scaffolding and lifts to small accessories. Ramirent provides services for the construction companies, subcontractors, installation companies, industrial plants, shipyards, local and country authorities as well as individual clients.

For a company which boasts the prestigious title of “Rental Company of the Year” granted by the industry-recognized European Rental Association, professionalism and maintenance of the highest customer service standards is a priority. To do this, both at the central and local level, actions are taken aimed at the improvement of services and their quality as well as the improvement of business continuity and safety. Also in the IT area, works aimed at ensuring the uninterrupted operation of IT systems even in the event of large natural disasters are carried out.

Ramirent came to the conclusion that it was necessary to use a backup data center, within the IT infrastructure in Poland, that would be located far enough away to quickly take over the tasks of the main data center in the event of a serious, unexpected failure.

On a company scale, it was one of the most important IT projects in 2010. SNP data center hosts the backup infrastructure. The SNP Outsourcing Center provides the adequate competence to ensure the maintenance and monitoring of the backup environment.

Backup IT

The main data center of Ramirent manages all the company’s IT systems, including user profiles used remotely from locations distributed across the country. The critical IT systems are run in a high-availability architecture (Microsoft Cluster Service). In the case of an unexpected failure, the data center has backup connections as well as its own power generator able to sustain the operation of servers for a pre-determined time. The standards met by the processing center are compliant with the best practices in IT management.

Ensuring a business continuity service in the event of serious disasters is not possible without a backup data center and a configuration of systems that will be possible to switch to in a pre-determined and appropriately short period of time. Moreover, the backup systems should provide services at a level that the users are accustomed to – both as regards the response time and availability.

Kamil Lisiecki, IT Manager in Ramirent, Poland talks about the direct reasons for the decision: “Security and the IT business continuity is our priority. With the distribution of our branches and employees, the maintenance of a reliable communication with the company’s systems requires additional protection in order to keep the restrictive availability parameters required by the business.

However, we don’t have an influence over everything. The best example is a huge power outage in Szczecin that took place two years ago. Furthermore, we are located in a flood-prone region – irrespective of the protection level, the servers won’t win against the elements. Even if our infrastructure were to survive, we would have to reckon with the fact that in such events the Internet providers or the electrical power providers are not able to guarantee a failure-free operation of their services. That’s why we have decided to launch a backup data center in an independent, external location. A continuous replication of the business data ensures a high level of security.”

Needed and may they not be useful…

Due to the high costs of building a second data center, Ramirent has decided to use the SNP data center as its backup center. In effect, it received a guarantee that key IT systems will be in operation and at a reasonable price.

The contract with SNP guarantees that the “backup” systems and data will be made available to employees and customers within half an of the failure notification. That’s an acceptable time from a business perspective. In practice, this means that in the event of a serious breakdown, one that would render the operations at the center in Szczecin impossible, the backup center would start to operate. Thus, the users may still carry out their everyday tasks. It is worth noting that switching to a backup installation is only expected in very exceptional circumstances.

Data replication

Within the backup data center service, the SNP Outsourcing Center provides Ramirent with a backup infrastructure collocation service along with hosting for the supporting systems which are required for the maintenance of the environment. SNP is responsible not only for the maintenance of the backup data resources, but also it has created the infrastructure which is necessary for the data replication to take place between the two installations – the main center in Szczecin and the SNP backup center.

The most important services that have been covered by the replication to the backup system include: the integrated ERP system, Microsoft SQL Server database, Microsoft Terminal Services, the Distributed File System and the Domain Name System, (DNS). Additionally, the user accounts in the Active Directory are also replicated.

Replication between sibling databases is designed to minimise the time that the main systems are unavailable in the event that the main data center in Ramirent is rendered inoperable. The database in the SNP data center remains in a standby mode, waiting to begin work in the event of a failure.

To improve database performance and accelerate its replication, SNP consultants performed an additional reorganization of the tasks inside the database, including the division of the data into what are referred to as transfer packages.

The data replication is done through a secure encoded WAN connection in a site-to-site configuration. The throughput of this channel depends on the number and size of the transferred data packages and customer expectations as to how long they are willing to accept the system’s unavailability in the event of a failure.

Data replication monitoring

To ensure the service continuity, the SNP Outsourcing Center continuously monitors the resources of Ramirent that are available in the SNP data center. The monitoring includes: a data connection layer, servers, systems and services, and is done in real time (online). Also examined are the individual processes and tasks responsible for the entire replication.

In the SNP Outsourcing Center, the monitoring service is provided based on the business continuity plans in the event of a failure and the certified security procedures, among others, in accordance with the ISO 9001:2008, ISO/IEC 20000 and ISO/IEC 27001 standards. A fully automated monitoring system consists of several integrated independent applications. Consultants are available in 24 a day, 7 days a week.

Business continuity procedure

Within the launch of the replication service, a business continuity procedure (BCP) that was created and tested in order to make a smooth transition from the main data center to a backup data center. It is a component of the IT business continuity procedure in Ramirent. This document sets forth in detail the actions to be taken in the event of a failure and other non-standard incidents, the steps to be taken in response to the well-documented, unusual situations, including, of course, the procedure for switching to a backup center. The tests are executed once a year and check the accuracy of the transition, the time needed for full recovery and the impact on the on-going work of the systems.

The business continuity procedure also sets up the rules for the on-going working relationship: the creation of backups in a backup center, etc. to ensure the data and system security in this location.

The system administrators both from SNP and Ramirent are obliged to familiarise themselves with this procedure.

What if…

At this moment, every few minutes, another batch of data from Ramirent’s IT systems is being sent to a backup system in the SNP data center. This process is run in the background and is not noticeable by the application users. Monitoring the process continuously ensures an instant response to any deviations (exceeding threshold values) which would necessitate taking certain corrective actions.

The security rules and the high-availability IT system implemented in Ramirent practically make it impossible that the unexpected failure of a single server, a main Internet connection or a short power outage would cause the main data center to fail.

A backup center is provided as insurance in the event of huge natural disasters (e.g. flood) or other unforeseeable events (fire, acts of terror, a large scale Internet connection failure). If such a critical event were to take place, leading to IT systems becoming unavailable during work, the IT Manager on behalf of Ramirent shall decide whether to switch. At the same time, the business continuity procedures will be instituted by the SNP Outsourcing Center.

As a result, within a certain time, a backup data center in SNP will take over the function of the main data center. All employees will receive a message about the failure with the instructions to use “backup” connections that will enable logging into the systems.

For the users who will work in the new infrastructure as of the switch, this change will be imperceptible. Their authorization scope or system interfaces will not change. Also, the system performance will be ensured at a proper level.

Within the business continuity procedure, the rules of conduct and liabilities in the recovery process of the main data center have also been prepared. A relevant procedure has been tested within the project.

Just in case

As Kamil Lisiecki concludes: “Thanks to the use of the solutions proposed by SNP based on the database server replication in a standby mode, we can save a lot on licenses for the second database. This was an additional argument. However, if we consider the backup data center service in terms of finances, I recommend paying more attention not to the costs of its on-going maintenance, but to the potential financial and business losses caused by the unavailability of the basic IT systems of our company. Therefore, I would compare a backup center to an insurance policy – we bear the costs of its maintenance and keep our fingers crossed for not having to use it.”

The SNP data center is located in the Nickel Technology Park Poznań (NTPP) in Złotniki near Poznan. The IT infrastructure meets the highest global requirements related to security and performance. The SNP data center in Złotniki is equipped with:

• Redundant power supply (two independent power lines), secured by redundant UPS systems operating simultaneously and additionally secured by a power generator
• Dual system fire detection and gas extinguishers for each zone in the data center
• Redundant air conditioning in the form of refrigerators
• Technical floor with moisture detection sensors installed underneath
• Motion detectors and seismic vibration sensors
• Fire and moisture detection systems integrated with the telephone exchange (in case of adverse events detection, the appropriate service is automatically notified)
• Monitoring (24 hours a day, 7 days a week) by a security agency and video cameras recording the picture
• Protected entry points that may be accessed by authorized personnel only.