Could you share the bucket name, the project URL, and roughly when the inconsistency started and ended? Also, were the differing reads happening from the S3 API, the dashboard, or both? That detail will help us pinpoint what went wrong.

Bucket: onepager-pulse-customer-data-hhes0c

https://railway.com/project/5d97afd6-47b3-41b8-b9d4-881cb71e81c0/bucket/b3b70f6a-f1d0-4e34-9630-ed48c48a02ce

6AM - 9AM PT

Issues in the Railway dashboard and via S3 API.

Facing a related issue: https://station.railway.com/questions/presigned-get-returns-no-such-key-for-obje-ac2c10af

Still waiting on a response here

We are looking into this. There is an underlying Tigris issue that may be the cause, but we will let you know more when we know more. https://status.tigrisdata.com/incidents/01KWFJYHP0MY6QQEWS87K6CYBB

The Tigris incident has been resolved. May I ask if you are still seeing this issue?

I have not seen the issue since Tigris resolved the incident.

Next time will Railway be publishing an incident in a timely manner to inform customers? This caused challenging situations with our customers, and we could not provide any timeline or information for resumuption of service.

Glad the issue cleared up with the Tigris incident resolution. Your point about timelier incident communication is valid - when an upstream issue like this impacts your ability to serve your own customers, faster visibility makes a real difference. That feedback is heard.