Fair use¶
Every account's automation runs on one shared BaseCloud platform. Fair use makes sure one account's automation volume can't slow down everyone else's. When your account runs a very high volume of automation at once, BaseCloud paces it: your tasks still run, just a little later than usual, until the volume drops.
Nothing is skipped or lost. Every task still runs, in order.
What you see¶
While any of a workflow's tasks is waiting its turn, a short note at the top of its canvas says:
Running slower than usual: fair use
Your automation throughput has been throttled under our fair use policy, because your account is running a high volume of automation right now. Nothing is skipped or lost: every task will still run, in order. It will just take a little longer than usual until the volume drops.
Each waiting task also shows an hourglass on the canvas. The task itself is still queued, just as it would be at any other time. The note is information, not an error. It goes away by itself as soon as the waiting tasks start.
Why fair use exists¶
Automation for every account shares the same workers and the same database. Without fair use, one account sending a huge batch, such as thousands of webhooks in a minute or an import that loops over every client, would put all of its work in front of everyone else's. Other accounts' messages, calls and forms would wait behind it.
With fair use:
- Capacity is shared fairly between accounts. While several accounts are busy, each gets an even share. When one account is quiet, the others can use what it isn't using, so no capacity sits idle while there is work to do.
- No single account can take the whole platform, however much it sends.
- Every account gets the same rule. There are no special cases and no plans that change it.
- Automation also gives way while the platform is under heavy load, including while many people are working in the CRM, so the screens you use stay responsive.
Nothing is skipped or lost¶
Fair use paces work. It never rejects it.
- Every task runs. A task that is kept waiting is not cancelled, skipped or dropped.
- Tasks run in order. Your workflow's logic is the same as it would be on a quiet day, just later.
- Waiting is not failing. A task waiting its turn keeps its place in line. It doesn't use up a retry, and it never shows as failed because it waited.
Your messages, calls and forms go first¶
Within your own account, your live work goes ahead of your bulk work:
- A Bulk Task is a Loop and every task inside its body, such as each step of an import or of a run over a long list of clients.
- A Live Task is every other task: the steps that answer a message or call coming in, a form being submitted or a webhook arriving.
So while a large import of yours is running, a new enquiry still gets its reply promptly instead of waiting for the import to finish. Your bulk work always keeps a share of its own as well, so it keeps moving and still finishes, even while live work never stops arriving.
BaseCloud decides which is which from the task itself. There is nothing to set.
Your workflows take turns¶
When several of your workflows are waiting at once, they take turns. One workflow receiving a large batch doesn't make your other workflows wait for the whole batch: each gets its turn in between, and each workflow's own tasks still run in order.
Large batches simply take longer¶
Pacing only happens while your account is running a high volume at once. On a quiet platform, tasks start the moment they are due, exactly as before.
When you send a very large batch, it takes longer to finish than it would if it had the platform to itself. That is all. Each task runs when its turn comes, and the batch finishes on its own.
Is there anything I need to do?¶
No. There is no setting to change, no limit to raise and nothing to contact support about.
You don't need to re-run a workflow or press Run Task again while the note is showing. The tasks are already in line, and they will run.
If you regularly send very large batches, you may notice they take a little longer at busy times. Your other automation keeps running promptly throughout.
Related¶
- Loop: how a Loop runs its body for each item
- Delay: waiting on purpose, for a time you choose
- Automation tasks overview