trigger
How it's scheduled
New appointments land continuously on the Batch / file feed. Front-of-house work is time-boxed — the estimate and letter have to reach the patient before the visit — so the pipeline polls every 30 minutes to process the visits within minutes of an appointment being booked, rather than waiting for an overnight batch.
| Scheduled task | Cadence | Server | State | Coordination |
|---|---|---|---|---|
\MDC\Daily - Processing - SSR | every 30 min (from 07:05) | prod-db-1 | enabled | standalone |
shape
Against the canonical Flow spine
Every Flow pipeline is a variation on one shape. Lit stages are the ones this pipeline runs; the labels match the capability lens.
task tree
What the pipeline runs
Tasks in order, each with what it does — typed stepsfrom the MDClarity task docs (click to open the reference), sub-jobsyou can expand inline (▶) to reveal their own tasks, and SQL steps summarized by Gemini (✨). Every step shows the tables it writes/reads and which earlier step it feeds from(sqlglot lineage).
Enqueue-Schedule-Data-Files
- DownloadDownloadDownloads files from remote FTP or SFTP servers to local directories with support for archiving and parallel transfers.remote directory=/local directory={SourceDataFolder}remote archive directory=StagingProcessingTest
- Delete Archived FilesExecuteProcesscommand=cmd.exearguments=/c del /F /Q /A "{SourceDataArchive}\*"
- Truncate File Queuesql
enqueue_schedule_data_files__truncate_file_queue.sqlwritesFileQueueScheduleData ▶Enqueue-Schedule-Data-Filesub-job · 1↗
- Load FileQueueScheduleDatasql
enqueue_schedule_data_file__load_filequeuescheduledata.sql✨ Records file names and paths into the schedule data queue for subsequent processing.writesFileQueueScheduleData
- FileQueue Checksql
enqueue_schedule_data_files__filequeue_check.sql· readsFileQueueScheduleData↳ feeds from Truncate File Queue