3 ms·
The inability to have other clients read data from S3 after S3QL has posted it here is a real dealbreaker for me. As an alternative I'd suggest S3FS - http://c
by michaelbuckbee 16y ago
The inability to have other clients read data from S3 after S3QL has posted it here is a real dealbreaker for me.
As an alternative I'd suggest S3FS - http://code.google.com/p/s3fs/ http://code.google.com/p/s3fs/
- StavrosK 16y agoIn their defense, the advantages this brings are really only possible the way they do it. I quite like this project, it just doesn't fit every use case.
- jevinskie 16y agoI used S3FS to upload 50k images (about 50KB each) and the performance was atrocious. The upload took an entire day on a 100mbit pipe and the s3fs userspace application ballooned to over 1GB of memory usage.
- asymptotic 16y agoDefinitely agree that a lack of interoperability with other S3 clients is a deal-breaker. S3FS made my day, thanks for bringing it to my attention. I'm looking forward to using it, particularly as it supports caching.
- crocowhile 16y agoS3FS is basically unusable on a linux machine: slow and prone to errors.