Comment 0 for bug 1810372

[Impact]

A user reports that cachefilesd will spin at 100% of a cpu when started on a filesystem where the free space is less than the bcull threshold and culling the cache is insufficient to free up space.

Investigation shows that this is because cachefilesd detects that culling is required, tries to cull, and does not realise that culling cannot free up enough space, so just keeps retrying.

[Testing/Reproducing]

I created a VM with a spare disk, mounted it to /raid (or whatever).
I changed the /etc/default/cachefilesd to start at boot, filled the /raid filesystem to over the bcull threshold, and started cachefilesd. When running top, the cachefilesd process is at the top using approx 100% of 1 CPU. It appears to be trying to free up space in the cache (that is not even used) because the filesystem is over the threshold for culling.

[Fix]

This is fixed upstream in 0.10.6:

* Wed Feb 3 2016 David Howells <email address hidden> 0.10.6-1
...
- Suspend culling when cache space is short and cache objects are pinned.

The particular patch is ce353f5b6b5b ("cachefilesd can spin when disk space is short.")

[Regression Potential]
The patch is small and easily understood and backports with minimal effort.