Introduction
Google Cloud offers a comprehensive suite of services and products designed to meet various computing needs, ranging from compute, storage, and databases to networking, AI, and security. Each service comes with defined Service Level Agreements (SLAs) that guarantee specific performance metrics, such as uptime, availability, durability, and latency. These SLAs ensure that businesses can rely on Google Cloud for consistent and reliable service, making it a trusted platform for mission-critical applications.
SLI (Service Level Indicator)
An SLI is a specific metric that measures the performance of a service. It quantifies how well a service is doing regarding a particular aspect, such as uptime, latency, or error rate.
- Example: Imagine you run a website, and you want to measure how fast your pages load. The SLI in this case might be the “percentage of page loads that complete in under 2 seconds.”
SLO (Service Level Objective)
An SLO is a target or goal for an SLI. It defines the acceptable level of service performance. SLOs are internal goals that you strive to meet to keep your service reliable and users satisfied.
- Example: Continuing with the website example, your SLO could be that “99% of all page loads should complete in under 2 seconds.” This is the target you set based on your SLI.
SLA (Service Level Agreement)
An SLA is a formal contract or agreement between a service provider and a customer. It specifies the level of service that the provider commits to delivering, often including penalties if these levels are not met. SLAs usually include SLOs but also cover other aspects like support response times.
- Example: If you’re providing a website hosting service, you might have an SLA with your clients that promises “99.9% uptime per month.” If you fail to meet this, you might be required to compensate the client, perhaps with a discount or credit for that month.
Real-Life Example
Let’s say you’re using a cloud service to host your website.
- SLI: You measure the uptime of the service, and find that it’s up 99.95% of the time over the past month.
- SLO: Internally, you have set a goal to maintain at least 99.9% uptime each month.
- SLA: The cloud provider has an SLA with you that guarantees 99.9% uptime. If they fail to meet this, they might offer you a refund or a service credit as compensation.
Summary
- SLI is the actual measurement (e.g., 99.95% uptime).
- SLO is the goal you set (e.g., 99.9% uptime).
- SLA is the formal agreement or guarantee (e.g., 99.9% uptime with penalties for failure).
GCP Services SLAs
| Service/Product | Uptime SLA | Availability SLA | Durability SLA | Latency SLA |
|---|---|---|---|---|
| Compute | ||||
| Compute Engine | 99.99% | 99.99% | – | N/A |
| App Engine (Standard) | 99.95% | 99.95% | – | – |
| App Engine (Flexible) | 99.95% | 99.95% | – | – |
| Kubernetes Engine | 99.95% for regional clusters | 99.95% | – | N/A |
| Cloud Functions | 99.95% | 99.95% | – | – |
| Cloud Run | 99.95% | 99.95% | – | – |
| Bare Metal Solution | 99.9% | 99.9% | – | – |
| Storage & Databases | ||||
| Cloud Storage (Multi-Regional) | 99.95% | 99.95% | 99.999999999% (11 9’s) | – |
| Cloud Storage (Regional) | 99.9% | 99.9% | 99.999999999% (11 9’s) | – |
| Persistent Disk | 99.99% | 99.99% | 99.999999999% (11 9’s) | Sub-millisecond latency |
| Cloud Bigtable | 99.95% | 99.95% | – | <10ms read/write (95th percentile) |
| Cloud Spanner | 99.999% (multi-region), 99.99% (regional) | 99.999% (multi-region), 99.99% (regional) | 99.999999999% (11 9’s) | 10ms for reads, 50ms for writes (99th percentile) |
| Cloud SQL | 99.95% | 99.95% | – | N/A |
| Cloud Datastore | 99.95% | 99.95% | 99.999999999% (11 9’s) | <10ms (99th percentile) |
| Cloud Firestore | 99.999% (multi-region), 99.99% (regional) | 99.999% (multi-region), 99.99% (regional) | 99.999999999% (11 9’s) | <10ms (99th percentile) |
| Filestore | 99.99% | 99.99% | – | N/A |
| Cloud Memorystore | 99.9% (for Redis) | 99.9% | – | Sub-millisecond latency |
| Cloud SQL | 99.95% | 99.95% | – | N/A |
| Cloud BigQuery | 99.99% | 99.99% | – | 3 seconds query execution (median) |
| Cloud Dataflow | 99.9% | 99.9% | – | – |
| Cloud Data Fusion | 99.9% | 99.9% | – | – |
| Cloud Composer | 99.9% | 99.9% | – | – |
| Cloud Data Catalog | 99.9% | 99.9% | – | – |
| Cloud Dataproc | 99.9% | 99.9% | – | – |
| Cloud Datalab | 99.9% | 99.9% | – | – |
| Firestore | 99.999% (multi-region), 99.99% (regional) | 99.999% (multi-region), 99.99% (regional) | 99.999999999% (11 9’s) | <10ms (99th percentile) |
| Networking | ||||
| Cloud Load Balancing | 99.99% | 99.99% | – | N/A |
| Cloud CDN | 99.95% | 99.95% | – | – |
| Cloud DNS | 100% availability | 100% availability | – | <10ms (global average) |
| Cloud Interconnect | 99.99% | 99.99% | – | 50-100ms (RTT latency) |
| Cloud VPN | 99.9% | 99.9% | – | – |
| Cloud Armor | 99.99% | 99.99% | – | N/A |
| Traffic Director | 99.99% | 99.99% | – | N/A |
| AI and Machine Learning | ||||
| AI Platform (Vertex AI) | 99.9% | 99.9% | – | – |
| Dialogflow | 99.9% | 99.9% | – | – |
| AutoML | 99.9% | 99.9% | – | – |
| Cloud Natural Language | 99.9% | 99.9% | – | – |
| Cloud Translation API | 99.9% | 99.9% | – | – |
| Cloud Vision | 99.9% | 99.9% | – | – |
| Cloud Speech-to-Text | 99.9% | 99.9% | – | – |
| Cloud Text-to-Speech | 99.9% | 99.9% | – | – |
| AI Hub | 99.9% | 99.9% | – | – |
| Operations | ||||
| Cloud Logging | 99.9% | 99.9% | – | – |
| Cloud Monitoring | 99.9% | 99.9% | – | – |
| Cloud Trace | 99.9% | 99.9% | – | – |
| Cloud Debugger | 99.9% | 99.9% | – | – |
| Cloud Profiler | 99.9% | 99.9% | – | – |
| Security | ||||
| Cloud IAM | 99.95% | 99.95% | – | – |
| Identity-Aware Proxy | 99.9% | 99.9% | – | – |
| Cloud Key Management Service (KMS) | 99.95% | 99.95% | – | – |
| Secret Manager | 99.95% | 99.95% | – | – |
| Access Transparency | 99.9% | 99.9% | – | – |
| API Management & Integration | ||||
| Apigee | 99.95% | 99.95% | – | – |
| Cloud Endpoints | 99.9% | 99.9% | – | – |
| Cloud Pub/Sub | 99.95% | 99.95% | – | <100ms (publish-to-acknowledge latency) |
| Cloud Tasks | 99.9% | 99.9% | – | – |
| Cloud Scheduler | 99.9% | 99.9% | – | – |
| Developer Tools | ||||
| Cloud Build | 99.95% | 99.95% | – | – |
| Cloud Source Repositories | 99.9% | 99.9% | – | – |
| Cloud Code | 99.9% | 99.9% | – | – |
| Cloud Deployment Manager | 99.9% | 99.9% | – | – |
| Cloud Container Registry | 99.95% | 99.95% | – | – |
| Hybrid and Multi-cloud | ||||
| Anthos | 99.9% for GKE on-prem | 99.9% for GKE on-prem | – | – |
| Traffic Director | 99.99% | 99.99% | – | N/A |
SLAs can change over time. Refer Google Cloud SLA Documentation for most up-to-date information.
Notes:
- Uptime SLA: This indicates the guaranteed uptime of the service. For example, a 99.9% SLA indicates that the service is expected to be available 99.9% of the time.
- Availability SLA: Specifies the availability percentage that Google guarantees for the service. (uptime + maintenance factors)
- Durability: Some services, particularly storage services, offer durability guarantees, often expressed in terms of “9’s” (e.g., 11 9’s for Cloud Storage Multi-Regional).
- Latency: Indicates the general latency metrics, where available.