Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/ShohjahonSohibov/repo-for-agent/llms.txt

Use this file to discover all available pages before exploring further.

UpdaterAgent relies on Hangfire to run all background work on a schedule. Jobs handle critical automation like importing loads, drivers, trucks, trailers, and brokers from TMS every three hours; syncing real-time vehicle positions from ELD providers every 25 minutes; sending Telegram notifications to drivers; and monitoring driver statuses. Without these recurring jobs, the system would require manual intervention for operations that need to run continuously and reliably across all tenant companies.

​
Hangfire setup

Hangfire is configured in src/UpdaterAgent.Api/Dependencies.cs using the Npgsql PostgreSQL storage provider. All Hangfire tables are created automatically in the background schema on first startup. Jobs are registered via JobsRegistrar.RegisterRecurringJobs() in Program.cs and only run outside the Testing environment.
The Hangfire dashboard is available at /hangfire. Access requires the admin role. Use it to monitor job history, requeue failed jobs, and inspect execution logs.

​
Job categories

​
TMS import jobs

These five jobs sync master data from the TMS/LogisticsCore system every three hours. Each job iterates all tenant companies and updates local records.
Job IDSchedulePurpose
tms-import:load0 */3 * * *Import loads from TMS
tms-import:brokers0 */3 * * *Import brokers from TMS
tms-import:drivers0 */3 * * *Import drivers from TMS
tms-import:trucks0 */3 * * *Import trucks from TMS
tms-import:trailers0 */3 * * *Import trailers from TMS

​
ELD position jobs

Position data from Zippy GPS is imported every 25 minutes and cleaned up daily at midnight to prevent database bloat.
Job IDSchedulePurpose
eld-import:positions*/25 * * * *Import current ELD positions from Zippy GPS
eld-clear:positions0 0 * * *Delete position records older than 72 hours

​
Notification jobs

These jobs keep drivers and dispatchers informed about load status and driver availability.
Job IDSchedulePurpose
send:load-info0 */1 * * *Send formatted load details to drivers via Telegram (pickup within 2 hours)
check:sleep-timer-expiry*/1 * * * *Check for expired driver sleep timers and notify dispatchers
The send:load-info job respects per-company notification interval settings stored in the BackgroundJob entity. It tracks LastRunnedAt per company to avoid sending duplicate messages.

​
QM webhook jobs

QuickManage webhook subscriptions expire and must be renewed periodically. A separate processor job handles incoming webhook events on demand.
Job IDSchedulePurpose
qm-webhook:renew-subscriptions0 0 */2 * *Renew all QM webhook subscriptions every 2 days
QM webhook processorOn-demandProcess load.created and load.updated events from QuickManage

​
Driver monitoring

Job IDSchedulePurpose
check:stationary-driver0 */3 * * *Identify drivers who have not moved during an active InTransit load

​
Required job attributes

Every recurring job in UpdaterAgent must carry two attributes: 
[DisableConcurrentExecution(timeoutInSeconds: 300)]
[AutomaticRetry(Attempts = 0)]
public async Task ExecuteAsync()
{
    // job body
}
DisableConcurrentExecution prevents the same job from running in parallel. The 300-second (5-minute) timeout acts as a deadlock safeguard — if a lock cannot be acquired within that window, the run is skipped. AutomaticRetry(Attempts = 0) disables Hangfire’s built-in retry mechanism entirely. Jobs that process data from external TMS or ELD systems can produce cascading errors and duplicate records if retried automatically. The system recovers naturally on the next scheduled run instead.

​
Multi-tenant job pattern

Every job iterates through all companies rather than running once per tenant. Before querying tenant-scoped data, each job sets the tenant context via ITenantContextService. A failure in one company’s iteration does not block processing for others. 
foreach (var company in companies)
{
    _tenantContextService.SetTenant(company.Id);
    // process data for this company
}

​
Manual triggers

Two endpoints allow on-demand execution without waiting for a scheduled run. Both are useful for testing or immediate re-processing. 
POST /api/eta-jobs/calculate
POST /api/eta-jobs/calculate-all
These are defined in src/UpdaterAgent.Api/Controllers/EtaBackgroundJobsController.cs.

​
Next steps

Job reference

Complete table of all 11 job IDs, CRON schedules, and detailed descriptions.

Integrations overview

Learn about the TMS, ELD, and webhook providers that these jobs interact with.

Build docs developers (and LLMs) love

Get started for freeTalk to us