# Hangfire.SqlServer

**Kind:** Service

**Source:** `Pams/API/Pams.API/Pams.API.csproj` (line 1)

**Part of:** [Pams](subsystem-pams)

NuGet package dependency

`Hangfire.SqlServer` is a NuGet package dependency in `Pams.API` that adds SQL Server-backed storage for Hangfire jobs. When configured by the API host, Hangfire stores queued work, job state, and worker coordination data in SQL Server.

## Diagram

```mermaid
sequenceDiagram
    participant Client
    participant API as Pams.API
    participant Hangfire
    participant SQL as SQL Server
    participant Worker as Hangfire Worker

    Client->>API: Submit work request
    API->>Hangfire: Enqueue background job
    Hangfire->>SQL: Store job and state
    Worker->>SQL: Poll queued jobs
    Worker->>Worker: Execute job handler
    Worker->>SQL: Update job state
```

## Usage

```ts
async function submitBackgroundWork(payload: Record<string, unknown>) {
  const response = await fetch("/api/jobs", {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
    },
    body: JSON.stringify(payload),
  });

  if (!response.ok) {
    throw new Error("The API could not queue the background job.");
  }

  return response.json();
}

await submitBackgroundWork({
  jobType: "process-record",
  recordId: "record-id",
});
```

## AI Coding Instructions

- Treat `Hangfire.SqlServer` as server-side infrastructure; client code should request work through API endpoints rather than access Hangfire storage directly.
- Keep Hangfire SQL Server configuration in the API startup and dependency-injection path that configures background job storage.
- Use job handlers with serializable arguments; avoid passing request objects, database contexts, or open connections into queued jobs.
- Check SQL Server connectivity and migration permissions when adding or changing Hangfire storage configuration.