What happened to Object Storage on 2/01/2024 to increase failures?
My product streams large files to around 50k customers and I utilize a few Linode S3 buckets for this purpose. When the HTTPS streams fail my product will fallover to AWS. Since 2/01/2024 those failure have increased 10 times.
- What changed at Linode?
- Are we penalized for large transfers?
- Will we ever get any kind of transfer statistics and failures reporting for our buckets?
Thanks in advance.
2 Replies
To my knowledge, we have not made any changes to our platform that would have affected customers' ability to stream from their OBJ buckets. When you say that your HTTPS streams have increasingly started failing after that point, what type of errors are you receiving, and are those errors persistent or intermittent?
Object Storage is factored into our overall Network Transfer allowance such that:
- Using OBJ grants you an additional 1TB/mo towards your global allowance pool.
- For every GB over your total global limit, you will be charged an additional $0.005-$0.015/GB based on the regional rates.
- Our transfer limits only incur additional charges and do not result in limiting your bandwidth capabilities.
Otherwise, although we currently do not have a platform-native monitoring tool, I have submitted a feature request on your behalf. I don't have a timeline of when this could be implemented, but be sure to follow our Blog for release updates:
I'm having intermittent issues where transfers return 503 Service Unavailable. I've recently installed a monitoring service that retreives a 2 byte file containing "OK" on all my object store buckets every minute and it even fails at least once a day. This tells me it has nothing to do with the size of the files. Is there anything support can do to investigate why I'm getting failures? What is the maximum limit of concurrent connections to a bucket?