Stability of the Workers infrastructure

Hey Guys!

Hope you are all well :slight_smile: I am running a frontend application hosted on CloudFlare Workers. We do deploy very often and for 3 days we are experiencing a production issue every day basically.

After the deployment, we are getting an error like

could not find index.61b26e3c3e.html in your content namespace

It takes some time until the site starts working again. Yesterday when that was happening I saw the issue with Workers update time (

I suspect that this issue has not really been resolved and it is still occuring. The workers’ platform looks to not be very stable. Can I ask Cloudflare staff to look into this issue more deeply and find the root cause? If this errors will occur I will have to implement a circuit breaker and serve the page from S3 and only utilize some CloudFlare caching.

Krzysztof Szarlej

I’ve seen the same, it seems the workers KV are evacuated prematurely on each update which cause a moment of KV’s not being found. Usually this is only a couple of seconds, but it might seem much more in a production environment, since this will occur on every cloudflare datacenter that runs the workers.

I remember that this isn’t a new issue, it’s been reported before but I’m unable to find the thread…

I notice that I get 404 errors after deploys every now and again (meaning I deploy a React app for example, and one of the scripts it includes can’t be found). I know there was an issue awhile back with Wrangler (where it would cause Workers KV (key/values) to be removed too quick), but I believe it was fixed. Interested to keep this topic bumped, so maybe we can get info from the Workers team to see if there are still issues :slight_smile:

However obviously this issue isn’t reproducible, as it only happens now and again (i’m guessing when certain POP’s are degraded or slower than others perhaps?)

1 Like

Are we talking about Workers Sites?


I believe so, yes :slight_smile: (Atleast I was in my post)