[Gluster-devel] Preventing lookups from serving metadata.
vbellur at redhat.com
Wed Nov 9 01:06:35 UTC 2016
On Tue, Nov 8, 2016 at 8:10 AM, Ravishankar N <ravishankar at redhat.com> wrote:
> So there is a class of bugs* in exposed in replicate volumes where if the
> only good copy of the file is down, we still end up serving stale data to
> the application because of caching
> in various layers outside gluster. In fuse, this can be mitigated by setting
> attribute and entry-timeout to zero so that the actual FOP (stat, read,
> write, etc) reaches AFR which will
> then fail it with EIO. But this does not work for NFS based clients.
Note that client side metadata caching in nfs can be disabled by
mounting with noac, lookupcache=none etc. However there would be a
performance impact by doing so.
More information about the Gluster-devel