# Size older workloads > Set resource requests on Orchard apps and databases that were created before resource limits existed. Source: https://docs.orchard.hackclub.com/operate/backfill/ The **Backfill** page lists apps and dedicated databases that have no size, which means no CPU or memory request. The scheduler reserves nothing for such a workload, so it can crowd out others on a busy node, and [resource limits](/operate/resource-limits/) cannot account for it. The page suggests a size for each one from its live usage, so you can size them all in one pass. When you apply a size, Orchard saves it and, if resource limits are on, restarts the pods with requests equal to limits. When resource limits are off, Orchard saves the size and leaves the running pods unchanged. ## Prerequisites - The superadmin role. ## Apply sizes 1. Go to **Instance admin** > **Backfill**. 2. Read the **Observed per pod** column. This column shows live usage, if metrics are available. 3. In **Apply size**, select a size for each row. To enter other values, select **Custom…**. To leave a row unchanged, select **Skip**. 4. Select **Apply** on the row, or select **Apply recommended to all**. Orchard recommends the smallest preset that fits 150% of the observed usage. A row without metrics has no recommendation and defaults to the smallest preset. When no rows remain, the page shows **Nothing to backfill**. ## Related resources - [Apply resource suggestions](/deploy/resource-suggestions/) - Resize an app later from its observed usage. - [Scale an app and set resources](/deploy/scaling/) - Set the size, replicas, and health checks of one app.