Getting random 404 from railway
I have a public endpoint https://api.hellowine.cl/api/v1/subscription_types, sometimes it's working fine but sometimes it's giving me 404 railway page.
< HTTP/2 404
< content-type: text/html
< x-railway-fallback: true
< content-length: 1596
< date: Wed, 15 Mar 2023 20:29:46 GMT
< server: railway
< x-envoy-upstream-service-time: 28
This is happening on every request, this public endpoint it's just an example.
5 Replies
Project ID:
270cce3a-27f2-4166-ab7a-4d1f07bed960270cce3a-27f2-4166-ab7a-4d1f07bed960
hello,
I've taken a quick look at this for you. I can see a few 40x responses in the last few hours. But our logs are showing that they came from the app.
does the response contain any body content? and was this happening in the 5 mins immediately following a new deploy?
If you can throw up a time window within which you see these responses I can try and drill down deeper.
Hello @char8 . This situation started suddenly without any deploy in between, so I panicked and started to launch a new deploy for every service.
I have a load-balancer service (haproxy) which has a 404 default rule for non existing routes, but this is not the case because its serves a known default haproxy 404 page. The response I was receiving was a 404 railway html page, the same when you request a non existing railway service or when you don't set the PORT env variable.
This was really random for whatever endpoint, to test this I was sending 10 request for the same endpoint and 2 or 3 of them gave the 404 html railway page.
By the time I'm writing this it seems the service its working good (no deploy in between), I'm not getting the random 404 html railway page anymore, but when we have the issue it affected a lot to our business 😢 .
Hi thanks for clarifying and sorry it affected your usage.
I took a look at logs between 00:00h UTC to 10:00am UTC today and I can't see any 404's from that endpoint on my side either. If this happens again, please let us know.