# A lag in using images from registry with a custom tag (latest)

**URL:** <https://community.fly.io/t/a-lag-in-using-images-from-registry-with-a-custom-tag-latest/26354>\
**Category:** Build debugging\
**Tags:** machines, registry\
**Created:** [October 30, 2025, 11:50pm UTC](https://community.fly.io/t/a-lag-in-using-images-from-registry-with-a-custom-tag-latest/26354 "2025-10-30T23:50:57Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![clintjhill](https://sea1.discourse-cdn.com/fly1/user_avatar/community.fly.io/clintjhill/32/43275_2.png) [@clintjhill](https://community.fly.io/u/clintjhill)\
**Post date:** [October 30, 2025, 11:50pm UTC](https://community.fly.io/t/a-lag-in-using-images-from-registry-with-a-custom-tag-latest/26354/1 "2025-10-30T23:50:57Z")

</div>

Is there an expected latency before images can be found in the registry using flyctl deploy?

I have an application that I use for a base image used by other applications in the same org. Every time after a brand new deployment, I encounter this error, that then magically goes away after some period of time. This has been problematic with my CI/CD that I’ve created and I can’t find issues in my own workflows, because after some period of time it just starts working with no other changes.

- deploy the app that is my base image with success.
- I can see the image using fly releases command.
- deploy another app that uses that base image and it fails the query.
- try again some time later and it works.

Is that expected behavior?

This is the error with an example name for the app:  
`failed to run query ($appName: String!, $imageRef: String!) { app(name: $appName) { id image(ref: $imageRef) { id digest ref compressedSize: compressedSizeFull } } }: Could not find image <``registry.fly.io/myappsname:latest``>`

---

<div class="post-metadata">

**Author:** ![clintjhill](https://sea1.discourse-cdn.com/fly1/user_avatar/community.fly.io/clintjhill/32/43275_2.png) [@clintjhill](https://community.fly.io/u/clintjhill)\
**Post date:** [October 31, 2025, 4:55pm UTC](https://community.fly.io/t/a-lag-in-using-images-from-registry-with-a-custom-tag-latest/26354/2 "2025-10-31T16:55:38Z")

</div>

Just a follow up on this, it works now and did not work just 10hrs ago. No changes on my part.

Is there a CDN-like cache priming that happens with [registry.fly.io](http://registry.fly.io)?

---

<div class="post-metadata">

**Author:** ![lubien](https://sea1.discourse-cdn.com/fly1/user_avatar/community.fly.io/lubien/32/41919_2.png) [@lubien](https://community.fly.io/u/lubien)\
**Post date:** [October 31, 2025, 6:12pm UTC](https://community.fly.io/t/a-lag-in-using-images-from-registry-with-a-custom-tag-latest/26354/3 "2025-10-31T18:12:07Z")

</div>

We do have some caching for tags which in turn makes `:latest` not always up to date. We’ve been using tags in some stuff we need to update instead of `latest` because of that.