This topic was automatically generated from Slack. You can find the original thread here.
We are planning to use the delay mode in one of our workflows. As per the pricing details, it is mentioned that charges are based entirely on compute usage.
To plan our budget effectively, we want to understand whether the duration of the delay in a workflow contributes to the compute credits consumed, or if delays are handled differently without incurring additional compute costs.
You should visit Pipedream for the latest pricing, but I’ll do my best to answer. The duration of a delay in a Pipedream workflow (using the Delay app or the $.flow.delay function) does not impact the number of compute credits consumed. Whether you delay your workflow for one second or several hours, the length of the delay itself does not increase credit usage. However, using a delay in a workflow will always incur two credits: - One credit is used when the workflow initially starts. - The second credit is used when the workflow resumes after the delay period ends. Example: If your workflow runs, hits a delay step, waits for an hour, and then resumes, you will be charged two credits for that execution (assuming default memory settings and no other special configuration). > Note: If your workflow’s execution timeout or memory configuration is higher than the default, credit usage may be higher. See the pricing docs for more details. Summary: - The duration of the delay does not increase credit usage. - Each delay incurs two credits (start + resume). For more information, see the Pipedream pricing documentation and workflow delay docs.
can you please verify this?
Pi’s info is correct, see these docs:
we have been charged like 48 credits for the given workflow after adding a delay and introducing a parallel execution?
What’s the memory setting on your workflow? That contributes to credit usage
If that’s not it, please open a support ticket with more detailed info so our support team can investigate it
