> For the complete documentation index, see [llms.txt](https://docs.pullbay.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.pullbay.com/documentation/platform-best-practices/reliability-metrics.md).

# Reliability Metrics

## Reliability Metrics

Pullbay tracks and publishes reliability metrics for transparency. Understanding these metrics helps you set expectations and design resilient integrations.

***

### Key Metrics

#### Availability

**Definition**\
The percentage of time the service accepts and processes requests.

**How it's measured**\
Availability = (successful checks / total checks) × 100

**What to expect**\
We target **99.9% availability** for production endpoints.

***

#### Success Rate

**Definition**\
The percentage of requests that complete successfully.

**How it's measured**\
(successful requests / total requests) × 100\
Client errors (4xx) are excluded from this calculation.

**Why it matters**\
Success rate reflects both Pullbay’s infrastructure reliability and the stability of upstream data sources.

***

#### Response Time

**Definition**\
The time from when a request is received to when a response is sent.

**Metrics tracked**

| Metric | Description                                                 |
| ------ | ----------------------------------------------------------- |
| p50    | Median response time (50% of requests are faster than this) |
| p95    | 95th percentile (95% of requests are faster than this)      |
| p99    | 99th percentile (only 1% of requests are slower)            |

**Why p95 and p99 matter**\
Average response time can hide slow outliers. p95 and p99 help you understand the performance of slower requests and design better timeouts.

***

### Metric Variations

Metrics can vary depending on several factors:

#### Endpoint Type

* **Standard pagination endpoints** — Faster per request (fetching one page)
* **Managed pagination endpoints** — Slower overall (fetching multiple pages internally)

***

#### Data Source

Different upstream platforms have different reliability profiles. Some are more stable, while others may change more frequently.

***

#### Time of Day

Some data sources may experience higher load at certain times, which can affect response times.

***

### Viewing Metrics

#### Per-endpoint Metrics

Your dashboard provides metrics for each endpoint you use, typically across:

* Last 24 hours
* Last 7 days
* Last 30 days

***

#### Historical Data

Historical metrics help you:

* Identify usage patterns (e.g., weekly cycles)
* Detect performance degradation trends
* Plan capacity and scaling decisions

***

### Setting Expectations

When building integrations, consider the following:

| Factor          | Recommendation                                     |
| --------------- | -------------------------------------------------- |
| Timeout         | Set to at least **2× the p95 response time**       |
| Retry count     | **3 retries** is usually sufficient                |
| Circuit breaker | Trigger after **50% failure rate over 30 seconds** |

***

### What We Don’t Guarantee

#### Data Freshness

Pullbay fetches data from upstream sources. Some data may be cached or delayed.

***

#### Data Completeness

Upstream platforms may limit the amount of data available. Responses may not always include the full dataset.

***

#### Schema Stability

While we aim to keep schemas stable, upstream changes may require updates to response structures.

***

### Summary

Reliability metrics help you understand how Pullbay performs in real-world conditions. By using these metrics to guide your timeout, retry, and error-handling strategies, you can build more stable and resilient integrations.