I'd like an Agent to check a few websites at regular intervals and collect the updates in a spreadsheet.
One thing I'm not sure about: if a website is slow and the task is still running when the next run is due, does Buda skip that run, queue it, or start another one?
I'm a bit worried about two runs writing to the same spreadsheet at once. Is there a recommended way to set this up?
It depends on the session mode in Buda's scheduled task. With a new session for each run, the next scheduled tick creates another task and the two runs may overlap. With a reused, pinned session, a tick that arrives while that session is still pending or running is queued for the same session. If the pinned session is waiting for your input, the automatic tick is recorded as skipped instead. A manual Run now against a busy pinned session reports that the previous run is still in progress.
For your spreadsheet, I would choose a reused session and a schedule longer than the usual website-check duration. Also make writes idempotent: use a stable key such as site + page URL + observed date/version, so a retry updates the same candidate instead of adding a duplicate. Reusing one session avoids two runs of this particular schedule writing simultaneously; it does not prevent another agent or workflow from editing the same spreadsheet. Watch the run history for running, skipped and failed entries, and decide how stale observations should be handled.
Is it more important that every scheduled interval eventually runs, or that overdue intervals be dropped rather than writing late results? The Chinese discussion covers the same two modes.
Sign in to join the discussion. Reading is always open; you only need an account to post.
Sign inI'd like an Agent to check a few websites at regular intervals and collect the updates in a spreadsheet.
One thing I'm not sure about: if a website is slow and the task is still running when the next run is due, does Buda skip that run, queue it, or start another one?
I'm a bit worried about two runs writing to the same spreadsheet at once. Is there a recommended way to set this up?
It depends on the session mode in Buda's scheduled task. With a new session for each run, the next scheduled tick creates another task and the two runs may overlap. With a reused, pinned session, a tick that arrives while that session is still pending or running is queued for the same session. If the pinned session is waiting for your input, the automatic tick is recorded as skipped instead. A manual Run now against a busy pinned session reports that the previous run is still in progress.
For your spreadsheet, I would choose a reused session and a schedule longer than the usual website-check duration. Also make writes idempotent: use a stable key such as site + page URL + observed date/version, so a retry updates the same candidate instead of adding a duplicate. Reusing one session avoids two runs of this particular schedule writing simultaneously; it does not prevent another agent or workflow from editing the same spreadsheet. Watch the run history for running, skipped and failed entries, and decide how stale observations should be handled.
Is it more important that every scheduled interval eventually runs, or that overdue intervals be dropped rather than writing late results? The Chinese discussion covers the same two modes.
Sign in to join the discussion. Reading is always open; you only need an account to post.
Sign in