Noah Watkins
Noah Watkins
@szym yes, we are also seeing this Arrow
> Or is the only reason you want a different allocation behavior because you need smaller chunks Correct. We operate in an environment where memory fragmentation exists. After the system...
> I have created a PR implementing that here: #112 > > Let us know what you think. 👋 👋 👋
Should we manually trigger the github action? I wonder if we can configure the action to auto-build if the action itself is changed and override the cron settings.
> Something to consider here? https://buildkite.com/redpanda/redpanda/builds/48328#018f1ab6-8870-4a42-a6a9-9b011310d2fc what am i looking at here? smokecheck?
/ci-repeat 1
@andijcr can you take a look at the ci failures?
what's the status of this @andijcr? its approved, but it looks like something changed? (the re-request review buttons are active).
I'm not sure how they used create() to do that either. I think we should use ioctl, though either way. https://github.com/libfuse/libfuse/blob/master/include/fuse_lowlevel.h#L868 I've never tried to use ioctl with fuse so...
Sweet... looks like there is an example of Ceph of them using ioctl in FUSE for reference https://github.com/ceph/ceph/blob/master/src/client/fuse_ll.cc#L498