For all mainnet deployments, Caldera offers a 99.95% uptime SLA.
Industry Standard Best Practices
At Caldera, we employ industry-standard best practices to ensure the reliability and availability of our services. Our Kubernetes configurations are designed with a high availability setup to mitigate service disruptions and maintain a resilient infrastructure. By adhering to recognized standards and best practices, we are able to provide a robust and reliable environment for our operations and, by extension, our partners.
All Caldera infrastructure is hosted in Amazon Web Services, and split across multiple regions. Our Kubernetes configuration is designed with reliability in mind: we auto-failover in the case of sequencers going down, and we autoscale our nodes when traffic spikes.
L1 RPC Fallback
One of the “weak links” in rollup operation is the connection to a layer-one (Ethereum mainnet) full node. If this connection is severed for an extended period of time, the rollup cannot function properly.
To further bolster our reliability, we have established a proxy load-balancing service to derisk potential challenges posed by L1 RPC outages. This proxy service aggregates over multiple top RPC providers, and re-routes our traffic in the event of downtime. This layer of protection helps in maintaining uninterrupted service and ensuring that our systems remain accessible and functional even during external disruptions in the broader network.
Transparent Status Page
We believe in full transparency when it comes to system performance and availability. Every mainnet deployment comes with a status page, powered by Betterstack. Our status page provides real-time updates on each rollup, allowing you to subscribe and stay informed about the operational status of our services.
In the event of downtime, all members of the Caldera engineering team are alerted.
Response SLAs and Escalation Procedures
Timely response and resolution are crucial for maintaining a high level of service reliability.
We have defined Service Level Agreements (SLAs) to ensure prompt responses to incidents and issues, 24 hours a day, 7 days a week. Our SLAs outline the expected response times and resolution procedures, providing a clear framework for addressing concerns. Additionally, we have an escalation procedure in place for operational and arbitration scenarios, ensuring that critical issues receive the necessary attention and are resolved expediently.