Synching is really slow about riofs HOT 2 CLOSED

The main purpose of RioFS is to act as a local filesystem, with S3 as "backend".
Using this idea we created an optimized for local operations filesystem, trying to minimize delays and caching as much data locally as possible.
Achieving realtime conformity between local filesystem and remote S3 bucket never been the objective of RioFS.

When you first get the list of files, RioFS sends a request to S3 and stores the list of received objects in the local cache. Then further "directory listing" operations use cached local list of files / directories, without sending requests to S3.

If you want to have a "close to realtime" version of S3 filesystem, you could try to set caching timers to a low value, please check dir_cache_max_time variable in riofs.conf.xml configuration file.

But please note, that I've never really tested RioFS with caching disabled.

from riofs.

Oh, I just approximated that the cache time was 5 minutes. I didn't know that it was actually five minutes. Thanks. I'll set this value lower.

from riofs.