Scheduled maintenance and downtime on October 15

Hi everyone – I wanted to give you a heads up about some important database upgrades we’re making to improve Airtable’s reliability. Our services will be unavailable for about 20 minutes on Friday, October 15 at 10 p.m. PT, and we wanted to give you a heads up given the potential disruption to your operations. Here are the details, which can also be found here in our help center.

We’re making important database upgrades to improve Airtable’s reliability. On Friday, October 15 at 10 p.m. PT, our services will be unavailable for about 20 minutes as we complete this important work.

Although these upgrades will improve Airtable’s performance, speed, and ability to scale further in the future, we recognize that any downtime is disruptive to our customers. We’ve identified this timing based on historical usage patterns and the fact we’re able to avoid working hours for the vast majority of our users.

While we believe this necessary maintenance will be complete in only a few minutes based on our testing and prior experience, we are accounting for up to 20 minutes of downtime. This is the first scheduled maintenance we’ve conducted since January 2020, as we never want to take our services offline unless it’s strictly necessary.

Thank you for your patience! We will continue to keep you informed as we complete this important work.


We’ve received a few questions about the scheduled maintenance, so I wanted to share some more details for those who are interested.

How can I receive updates about the scheduled maintenance?

You can subscribe to updates here on our status page.

How are API requests affected?

During the maintenance window, all customer API requests will return 503 responses. These requests should be retried with exponential backoff.

How will automations be affected?

Automations will be disabled for the duration of the maintenance window, and queued jobs will be paused during maintenance.

How will Airtable Syncs be affected?

Depending on the exact moment they happen, trigger events from external sources (e.g. Google Calendar) may result in an error. If an error occurs, the sync will attempt to re-run (when and how often depends on the sender). This means that in some cases synced data may fail.

How will time-based formulas be affected?

We have fallback logic built in to retry time-based formulas in the event that they don’t update.