Trusted Execution Environment Cloning via Escrow Snapshots
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
FaaS workloads in trusted execution environments face inefficiencies due to the high cost of rebuilding workload containers from scratch for each service request, as current technologies do not allow replication of trusted execution environments from checkpoints, leading to significant overhead in TEE creation and state copying.
Innovation Solution
A system with an escrow, snapshot, and cloning service trust domain capability enables the replication of a hibernated state of a trust domain or secure enclave, using hardware-measured and sealed escrow and cloning policies to instantiate TEE clones locally or remotely without a virtual machine manager in the trusted computing base, allowing baseline TDs/enclaves to be reused for function execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If FaaS workloads are rebuilt from scratch for each service request in TEE, then security and isolation are maintained, but operational efficiency deteriorates due to high overhead in TEE creation and state copying
Solution Approach 1:
The patent implements workload replication by creating copies of TEE workloads from checkpoints. The system captures the state of a TEE workload at a checkpoint and replicates it to multiple target platforms, allowing the same workload to serve multiple service requests without rebuilding from scratch. This copying mechanism maintains security through hardware-enforced isolation while dramatically improving operational efficiency by eliminating repeated TEE creation overhead.
Solution Approach 2:
The patent employs checkpointing as a preliminary action where the TEE workload state is captured and preserved before being torn down. This checkpoint serves as a pre-prepared template that can be quickly replicated and restored on demand. The preliminary capture of workload state eliminates the need to rebuild the entire TEE environment for each new service request, significantly reducing startup time and resource overhead while maintaining security boundaries.
2Quantity of substance
If TEE workloads are torn down after service requests, then resource utilization improves, but startup time increases for subsequent requests
Solution Approach 1:
Instead of rebuilding TEE workloads from scratch after teardown, the system creates replicas from preserved checkpoints. The original workload is torn down to reclaim resources, but its checkpointed state is replicated to target platforms when needed. This copying approach allows efficient resource reclamation while maintaining quick workload restoration capability, resolving the trade-off between resource utilization and startup time.
Solution Approach 2:
The patent changes the state parameter of TEE workloads by introducing checkpointing and replication mechanisms. Rather than transitioning from active to torn-down state permanently, the workload state is preserved in a checkpointed form that can be rapidly restored. This parameter change enables the system to switch between resource-efficient teardown and fast restoration by controlling replication and restoration processes on demand.
3Productivity
If workload replication is enabled in TEE, then operational efficiency improves, but security and isolation are compromised
Solution Approach 1:
The patent introduces checkpoints as an intermediary mechanism between the original TEE workload and its replicas. The checkpoint captures the workload state in a secure, isolated manner and serves as a trusted intermediary for replication. This intermediary approach allows workload copying while maintaining security boundaries, as the checkpoint acts as a controlled interface that preserves isolation guarantees even as workloads are replicated across multiple platforms.
Solution Approach 2:
The system implements secure copying by replicating workload states from checkpoints to target platforms. The copying process is controlled and managed through the checkpoint mechanism, which ensures that replicas are created from authenticated, isolated source states. This controlled copying maintains security and isolation by preventing unauthorized access to the original TEE while enabling efficient workload replication for improved operational efficiency.
Data Source
AI summary
Scalable cloning and replication for trusted execution environments is described. An example of a computer-readable storage medium includes instructions for receiving a selection of a point to capture a snapshot of a baseline trust domain (TD) or secure enclave, the TD or secure enclave being associated with a trusted execution environment (TEE) of a processor utilized for processing of a workload; initiating cloning of the TD or secure enclave from a source platform to an escrow platform; generating an escrow key to export the snapshot to the escrow platform; and exporting a state of the TD or secure enclave to the escrow platform, the state being sealed with a sealing key.