High Availability for Public Transport: Centralized Control Room
Public Transport Control Rooms
A centralized control room is the nerve center of modern public transportation systems. The control room monitors entire metro lines, manages automated operations, and ensures passenger safety. For subway systems, automated metro lines, and bus networks, the control room must operate 24/7 without interruption.
RATP manages the Paris subway. It transports more than 10 million passengers every day. Line 1 of the Paris Metro is fully automated and driverless. This makes the role of the control room even more critical.
Why is the control room critical?
The operators of a metro line must react immediately to the slightest incident. An IT failure must never prevent operators from seeing the activity of the line.
The main tasks of the control room operators are:
- Monitoring the movements of trains on the tracks
- Handling incidents detected automatically by video cameras
- Managing communication systems and public announcements
- Overseeing safety operations and passenger information
A system failure in the control room directly impacts passenger safety and service continuity. All critical applications in the control room must be protected by high availability. Examples are gateways, databases, and web front-ends.
SafeKit high availability for public transport
RATP faced major constraints for the automation project of line 1:
- A tight deployment schedule
- Most applications were not originally designed for high availability
- A solution was needed for both Windows and Linux servers
- There was no budget for expensive hardware, such as SAN storage or network load balancing appliances
SafeKit was chosen for the following reasons:
- It is a purely software solution with no shared disks and no SAN
- It provides the same solution on Windows and Linux
- It delivers three critical functions: load balancing, automatic failover, and real-time data replication
- It is simple to implement on standard servers, easy to administer, and versatile
Real-world implementation at RATP
RATP deployed SafeKit with two types of clusters:
- Farm cluster: an active/active cluster with network load balancing and automatic failover
- Mirror cluster: an active/passive cluster with real-time data replication and automatic failover
Deployment approach
First, RATP defined the functional needs. Then, SafeKit was configured for each protected application. The applications themselves were not modified. SafeKit requires no physical change to the servers.
A SafeKit application cluster has:
- Two servers with the same operating system
- The application installed on both servers
- SafeKit installed on both servers
- One or more application modules configured for high availability and load balancing
Several application modules can run in parallel on the same cluster. This optimizes the use of server resources. RATP validated the SafeKit modules in three steps: testing, pre-production, and production.
Daily operations
IT operators now administer all application clusters with:
- The same high availability solution on Windows and Linux
- The same administration console for all clusters
- The same configuration and the same commands
- Unified monitoring for load balancing, high availability, and data replication
The backup servers of critical applications are located in a separate machine room. They run on separate power supplies. If an incident occurs, operations continue without loss of service for Paris subway passengers.
Key benefits
- Cost reduction: no expensive SAN infrastructure with shared disks and no network load balancing appliances
- Simplified administration: one console manages both Windows and Linux clusters
- High availability: load balancing, automatic failover, and real-time data replication
- Minimal project impact: existing applications require no modification
- Operational resilience: automatic recovery ensures continuous passenger service
🔍 Explore SafeKit
Find information about SafeKit features, architectures, videos, documentation, and the free trial.
| Resource | Description | Link |
|---|---|---|
| Key features | Learn why SafeKit is a simple and cost-effective high availability solution. | Discover the key features |
| Use cases | Learn how SafeKit keeps critical applications and infrastructure available. | Explore the use cases |
| Deployment model | SafeKit provides high availability without a SAN or shared storage. | Explore the deployment model |
| High availability strategies | Compare high availability for virtual machines with high availability for applications. | Compare the two strategies |
| Technical specifications | Check the technical requirements and limitations for SafeKit clusters. | Read the technical specifications |
| Videos | Watch demonstrations of SafeKit configuration and automatic failover. | Watch the technical videos |
| Architecture | Learn how the SafeKit Mirror cluster replicates files in real time and handles failover. | Explore the Mirror architecture |
| Architecture | Learn how the SafeKit Farm cluster balances network traffic and handles failover. | Explore the Farm architecture |
| Product comparison | Compare SafeKit with traditional high availability clusters. | Compare the solutions |
| Technical resources | Find technical documentation, download SafeKit, and access the free trial. | Access the documentation and free trial |
| Application modules | Use preconfigured SafeKit modules to set up high availability for your applications. | Explore the application modules |