|
Data Center & Disaster Recovery site Setups
IDBI Ltd has invested in a robust & scalable IT infrastructure to support its aggressive growth plans. As part of the process, the bank has set up an IDC level 3 Data Center & Disaster Recovery Site to ensure service availability & resiliency to meet the changing needs of the end users.
The salient features of the new Data Center & Disaster Recovery site are as follows:
Site Selection
The site selection for the Data Center & Disaster Recovery site is done so as to minimise the impact of natural disasters. Both sites viz. Belapur for DC & Chennai for DRS are physically separated by a large distance (they also lie in different seismic zones). This minimises the chances of both the sites getting affected by natural disasters like flood, earthquake etc at the same time.
Physical Safety & Security
- Both the sites are at the height of first & second floor. This protects the sites from local floods, water clogging & physical intrusion.
- Both the sites are constructed using 2 hours fire rated QED building blocks. The glass, wherever used for critical areas like server room, network room etc, is 2-hour fire rated. These features provide safety in preventing spread of fire to adjoining areas by creating isolated zones.
- Each area of the Data Center & Disaster Recovery site is monitored by High Sensitivity Smoke Detectors. Sensitive plenums, like area below the false flooring, area above the false ceiling at server room & network room. The fire detection systems are interfaced with Fire Extinguishing systems for automatic release of FM200. An integrated Building Management System interfaces & controls these various systems.
- Water leak detection systems are deployed at critical locations to detect any ingress of water into the Data Center. These systems are interfaced with BMS for giving audio-visual alarms to indicate the nature & location of problem.
- Ultrasonic rodent repellent systems are deployed in sufficient numbers to cover the complete area of the Data Center.
- Physical security is enhanced by the following measures:
- 24X7 deployment of armed security guards in & around the Data Center.
- Each entrance/exit & cabin doors within the data center are under electronic surveillance by Closed Circuit TV cameras.
- Each door within the Data Center is a 2-hour fire rated metal door & is operated by proximity card/pin based electronic access control.
- The entry to sensitive areas like server room is controlled by biometric access control.
- All areas & sensitive equipments like diesel generator, UPS, Precision air conditioners, etc, are under 24X7 monitoring for physical access.
- Single point entry & exit from the data center with computerised visitor management system that stores the photographs of the visitors.
- Highly restricted entry for outsiders to the high sensitivity/high security area of the Data Center. Provisions are made for meeting rooms outside the high security area for external visitors.
- Material movement is restricted across the high security area by providing locker room facility outside the high security area for persons deployed at the data center.
- Access to various internal areas like Server room, Network room, Console room, etc, is given on “Need to Access” basis. No free movement is permitted within the high security zone.
Logical Security
Logical security of the Data Center is enhanced by the following measures:
- Multiple levels of the firewall are deployed at each Internet & extranet links.
- Physical access to the server is not permitted. All servers are accessed through KVM links provided in a secured Console room.
- Access to the Console room & to the servers through the KVM is on “Need to Access” basis. Audit trails are maintained for each server access & activity.
- Any changes to the application, OS or Database is controlled through a workflow based ‘Change Request Form (CRF)’. The CRFs are validated or approved by competent authorities.
- Multiple copies of critical data are maintained within a high availability SAN storage. The backups of critical data are taken on tapes & the tapes are stored at remote locations in fire rated cabinets.
High Availability Features
Critical components of the physical infrastructure & the compute infrastructure are deployed in redundant configurations to ensure maximum uptime of the end-user services. The high availability features of the DC & DRS are as follows:
-
Dual Rail Power Supply :
Both the data centers are provided with dual rail power supply system. Incoming power is tapped from two independent transformers & fed to two independent UPS. Power distribution till the server racks happen through independent power distribution panels. Each rack is provided with two power distribution strips powered by two independent UPS. Thus, there is no common point between the two power sources & each distribution system is capable of working independently & providing the complete power requirement of the Data Center. This ensures a complete power system redundancy & high uptimes.
-
Redundant Network Configuration :
The DC & DRS are pre-wired for full capacity. Each rack is provided with redundant network connections, which use separate layer2 & Layer 3 switches. This ensures the availability of the services even if a network component fails.
-
Climate Control :
Every critical area of the Data Center like Server room, Network room, etc is covered by independent precision air conditioning systems. Here the temperature & humidity is controlled with a high degree of precision. Each PAC system for individual area is configured with redundant units so as to avoid any “Single Point of Failure”.
-
Enterprise Class Compute Infrastructure :
All the servers hosting critical business applications are with built-in high availability features like redundant power supply units, redundant network & SAN connectivity, and multiple CPU/Memory boards. This minimises the risk of service outages arising out of any hardware failure of the critical servers/storage.
-
Cluster Deployments :
All business critical applications, like Core banking, ATM switch, etc are deployed on clustered server configurations. This minimises the application downtime due to server hardware failure.
IT Service Management
As part of the ITSM, IDBI has deployed Tivoli suites to effectively manage the IT infrastructure. Tivoli product range is being used for managing and monitoring the Network, OS monitoring, DB monitoring, deployment of patches and Storage Management. The suites are configured to send event alerts directly to the team looking after the infrastructure. This helps in reducing the time required to address any infrastructure related issues. Remedy is used to keep a track of the IT assets and Change Management.
IDBI is in the final stages of installing ‘Panaces’, the DR management software of Sanovi Technologies Ltd to monitor the RPO (Recovery Point Objective) and RTO (Recovery Time Objective) objectives envisaged by the Bank. This software would help in maintaining proper control and consistency of the application and database. IBM Global Mirror (PPRC) and Sanovi File replicator is being used to copy the volumes and files to the DR site.
|
