Bangalore International Airport Limited (BIAL), known for its smooth operations powered by IT, is a prime example of IT infrastructure set up and management. Reputed to be one of the best operated airports in the world, BIAL has now incorporated its disaster recovery implementation which has a key role to play in the airport’s business.
BIAL has its primary data center and disaster recovery (DR) site at the airport itself. There are about 60 servers of a heterogeneous nature (varying from Windows to Unix to Linux) that host different applications. These servers are spread across two data centers. The servers are mostly HP Proliant, HP 9000, HP Integrity, Dell, and blade servers from IBM.
The disaster recovery implementation’s reach
The Bengaluru airport’s disaster recovery implementation is hosted in a different building, but within the same campus. As U Nedunchezhiyan, the senior manager of ICT (Infrastructure) at Bengaluru International Airport points out, “Disaster recovery implementation as a business requirement is such that we cannot have the DR site far from the airport. If there is an issue at the Bengaluru airport and the DR site is far away, it is not possible to immediately start it.”
The data center has standard BSES power and UPS backup (with N+1 redundancy and regular battery backups). In terms of connectivity, telecom providers have terminated fiber connectivity at BIAL’s fiber head over-point. From this point it goes to BIAL’s telecom center and then to the users. Connectivity is not available directly between the users and telecom service providers. Apart from having its own infrastructure, BIAL also acts as service provider to many tenants within the airport. BIAL also provides a common IT infrastructure for all airlines.
The entire Bengaluru airport campus is being managed by fiber connectivity with redundant loops. As part of the disaster recovery implementation, the airport has ensured network and service redundancy. Arun S, the senior manager of ICT systems for BIAL says, “Our IT needs are completely different from that of other verticals. Even a second’s difference means a huge loss for us.” That’s why BIAL’s disaster recovery implementation is considered to be very critical by the airport authorities.
BIAL’s DR setup is very closely linked to its recent storage virtualization deployment. Currently, BIAL has gone for Hitachi Data Systems for virtualization of its storage infrastructure. Earlier, it had been using HP EVA (Enterprise Virtual Array) storage.
Critical airport applications
Replication for different applications is performed in a different manner in BIAL’s disaster recovery implementation strategy. Business critical applications are protected using synchronous replication. Applications critical for BIAL’s functioning include those such as the airport’s database—the repository for airport operations, ERP and middleware. BIAL’s DR site has replication of all business critical applications. However, it does not perform exact replication of the primary site.
Security plays a crucial part in BIAL’s disaster recovery implementation plan. It begins with physical security. The data center is protected by biometrics access and digital access. Each room is programmed such that only the concerned person has access. There is an in-house audit team that periodically checks for vulnerabilities. BIAL follows the standard ITIL process.
Work on recovery time objective (RTO) and recovery point objective (RPO) is underway at the moment as part of the disaster recovery implementation. Nedunchezhiyan says, “Right now there is tape-based back-up. We are planning to move to disk-based backups in order to improve RTO and RPO.”
The airport has stringent control on downtime since domestic flights dominate the day hours, and a large number of international flights operate at night. As a result, permission is required from every group for any sort of disaster recovery implementation. With the entry of GVK, managing airports scenario might change considerably. As Arun mentions, the IT team might have to rework the entire IT DR strategy with DR site in a different place. But at the moment, that migration is a long way off for BIAL.