[Linux-cachefs] Kernel traceback on 3.10.10

[Mark Moseley]

On Wed, Sep 11, 2013 at 9:16 AM, Milosz Tanski <milosz at adfin.com> wrote:

> Mark,
>
> Is a read only work load or a combination of a read / write workload?
>
> -- Milosz
>
> On Fri, Sep 6, 2013 at 7:52 PM, Mark Moseley <moseleymark at gmail.com>
> wrote:
> > Was testing out fscache for the first time in a while. On a fairly busy
> > box, I got this after about an hour of traffic. OS is Ubuntu Precise
> > 64-bit, on a Dell Poweredge 420. The cache is on a 500gb SATA drive:
>


This is a read-write workload. I used to test on solely read-only
environments (not that the data was read-only, just that writes were
happening on other servers) but we've since consolidated things a bit. It's
web hosting, so it's a good deal of ops/sec and can be an unpredictable
workload. It could've been talking to either Netapp or Oracle NFS gear,
though no way to tell which it was talking to when it popped.