> For the complete documentation index, see [llms.txt](https://docs.pharos.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.pharos.ai/open-telemetry/understanding-otel/otel-and-pharos.md).

# OpenTelemetry & Pharos

How Triton and Hermes divide the work of getting Salesforce telemetry into OpenTelemetry, and why Pharos never sees your exported data.

Pharos participates in OpenTelemetry through two components with a clean division of labor: **Triton** instruments Salesforce, and **Hermes** exports what Triton — and Salesforce itself — produces.

***

## Triton: instrumentation and identity

Triton is Pharos's instrumentation framework for Salesforce. It adds explicit logging calls across the runtimes a real transaction has to cross — LWC, Apex, and Flow — and propagates a single transaction identity through all of them: the `trace_id` an OTel backend later uses to reassemble the transaction. That shared identity is what lets a trace assembled downstream still read as one transaction instead of three disconnected fragments.

Triton itself is documented in full in the [Configuration Guide](https://docs.pharos.ai/configuration-guide/advanced/triton.md) and [Platform Overview](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/tracing/triton-tracing.md); this page covers only where it fits into the OTel picture, not the instrumentation APIs themselves.

***

## Hermes: OTel conversion and export

Hermes is the Salesforce → OpenTelemetry product family. It runs entirely in your own cloud, reads Pharos/Triton telemetry and Salesforce Event Monitoring, converts both into OTel spans, metrics, and logs, and exports them over OTLP to whichever OTel-compatible backend you already use.

Hermes doesn't need Triton to produce value on its own — Salesforce Event Monitoring telemetry flows through Hermes independently — but where Triton is present, Hermes carries its transaction identity into the exported trace, so the same transaction correlates end-to-end.

***

## The hybrid model

Pharos doesn't ask you to choose between Salesforce-native issue tracking and an external observability platform — it splits the work by volume and purpose:

| Stays in Salesforce                                                 | Flows to your OTel backend                               |
| ------------------------------------------------------------------- | -------------------------------------------------------- |
| Issues, error tracking, and the Slack/Jira workflows built on them  | High-frequency traces, metrics, and logs                 |
| Lower-volume, high-value signal your team already triages in Pharos | Detailed telemetry sized for a platform built to hold it |

You keep the issue-tracking workflow you rely on, while the telemetry volume that would otherwise strain Salesforce storage limits goes where it belongs.

***

## Where your data goes

Hermes runs in **your** cloud, not Pharos's. Pharos never receives, stores, or has access to the telemetry Hermes exports; Salesforce is treated as a read-only telemetry source, and the export path runs entirely between your Salesforce org and your chosen backend.

***

## How it fits together

```mermaid
graph LR
  A["Salesforce (Triton + Event Monitoring)"] --> B["Platform Events / EventLogFile"]
  B --> C["Hermes (your cloud)"]
  C -->|OTLP| D["Backend (Honeycomb, Splunk, ...)"]
```

Triton and Salesforce Event Monitoring produce the raw telemetry. Hermes reads it — via Platform Events for near-real-time signals, and EventLogFile for batch signals — converts it to OTel spans, metrics, and logs, and exports it over OTLP to your backend.

***

## Related

<table data-view="cards"><thead><tr><th></th><th></th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><strong>🔱 Configuration Guide: Triton</strong></td><td>Admin configuration steps for Triton</td><td><a href="https://docs.pharos.ai/configuration-guide/advanced/triton.md">Triton</a></td></tr><tr><td><strong>🔱 Triton Tracing</strong></td><td>Triton concepts and trace correlation</td><td><a href="https://docs.pharos.ai/platform-overview/pharos-platform-architecture/tracing/triton-tracing.md">Triton Tracing</a></td></tr><tr><td><strong>🛠️ Pharos Triton Instrumentation</strong></td><td>Triton logging APIs</td><td><a href="https://docs.pharos.ai/platform-overview/pharos-platform-architecture/logs/log-ingestion-methods/pharos-triton-instrumentation.md">Pharos Triton Instrumentation</a></td></tr><tr><td><strong>🔱 Triton API Reference</strong></td><td>Triton's own documentation site</td><td><a href="https://triton.pharos.ai">https://triton.pharos.ai</a></td></tr></tbody></table>
