3 months ago
Hey 👋
Ever since we moved one of our buckets over from AWS S3 to Railway Storage, we're noticing a very large increase in HTTP 408s that are coming directly from the storage endpoint, and we don't really understand why we're seeing so much of these.
Anyone else ran into the same before?
Note: I can't provide more specific info like bucket name or specific paths, as this application is very sensitive. I'm happy to give more information over DMs
8 Replies
Attachments
Note: I can't provide more specific info like bucket name or specific paths, as this application is very sensitive. I'm happy to give more information over DMs
fwiw these 408s usually happen in bursts with usually some time between them
Attachments
Just now I've updated our service to automatically retry these 408s we're getting from RW's Object Storage to see if that may help a bit
unfortunately retrying these 408s doesn't fix the issue. May try again by adding a little bit of a backoff period
but yeah still annoying and weird that there are very often blips where a lot of requests just don't come through and get answered with an 408. Would love to know why this is happening