[GH-ISSUE #1490] May support range cache? #786

New issue

Closed

opened 2026-03-04 01:48:47 +03:00 by kerem · 3 comments

kerem commented

2026-03-04 01:48:47 +03:00

Owner

Originally created by @fly3366 on GitHub (Dec 4, 2020).
Original GitHub issue: https://github.com/s3fs-fuse/s3fs-fuse/issues/1490

If anyone(like ffprobe) need only a few parts of large file. A parted cache(with file offset) may better. PR?

Originally created by @fly3366 on GitHub (Dec 4, 2020). Original GitHub issue: https://github.com/s3fs-fuse/s3fs-fuse/issues/1490 If anyone(like ffprobe) need only a few parts of large file. A parted cache(with file offset) may better. PR?

kerem

2026-03-04 01:48:47 +03:00

closed this issue
added the
need info
label

kerem commented

2026-03-04 01:48:48 +03:00

Author

Owner

@gaul commented on GitHub (Dec 4, 2020):

Doesn't this already work? What kind of behavior do you observe? You can see what s3fs is fetching via -f -o curldbg.

@gaul commented on GitHub (Dec 4, 2020): Doesn't this already work? What kind of behavior do you observe? You can see what s3fs is fetching via `-f -o curldbg`.

kerem commented

2026-03-04 01:48:48 +03:00

Author

Owner

@fly3366 commented on GitHub (Dec 8, 2020):

Sorry, My question:

Can bucketname.stat in cache dir shared between multi s3fs instance?
Can bucket in cache dir shared between multi s3fs instance?
Can s3fs cache object with offset? If I open file's offset like 0-1024K, May s3fs only request with range 0-102400 with curl? But not all file(0 - file end).

@fly3366 commented on GitHub (Dec 8, 2020): Sorry, My question: 1. Can `bucketname.stat` in cache dir shared between multi s3fs instance? 2. Can `bucket` in cache dir shared between multi s3fs instance? 3. Can s3fs cache object with offset? If I open file's offset like 0-1024K, May s3fs only request with range 0-102400 with curl? But not all file(0 - file end).

kerem commented

2026-03-04 01:48:48 +03:00

Author

Owner

@gaul commented on GitHub (Dec 8, 2020):

1&2) are not possible and will lead to data corruption
3) should work this way already. However s3fs will prefetch data equal to multipart_size * parallel_count.

@gaul commented on GitHub (Dec 8, 2020): 1&2) are not possible and will lead to data corruption 3) should work this way already. However s3fs will prefetch data equal to `multipart_size` * `parallel_count`.

kerem referenced this issue

2026-03-04 02:01:26 +03:00

[PR #786] [MERGED] Log messages for 5xx and 4xx HTTP response code #1639