Bug#700755: huge slab_unreclaimable in Xen domU

[Josip Rodin <joy@debbugs.entuzijast.net>]

On Wed, Feb 20, 2013 at 11:35:44AM +0000, Ian Campbell wrote:
> > OK, I can install both (it's got PV-GRUB), which do you prefer to test first?
> > I'm asking because it'll likely take a few weeks for the bug to appear,
> > judging by what it did before.
> 
> Probably at this stage I would be more interested in making sure Wheezy
> was going to be OK first.

ACK

> I'm not sure. The net_namespace one should be easy enough to track in
> the code since:
>         net_cachep = kmem_cache_create("net_namespace", sizeof(struct
>         net),
> and therefore users of net_cachep must be responsible, I'd expect there
> to be not all that many of those. Are you actually using network
> namespaces in the guest?

If I am, I'm not doing it intentionally :) I'd assume this was part of the
LXC functionality, but as Ben noticed before, there's code in vsftpd that
triggers their use...?

> The Debian kernels have SLUB:
>         /boot/config-2.6.32-5-xen-amd64:CONFIG_SLUB_DEBUG=y
>         /boot/config-2.6.32-5-xen-amd64:CONFIG_SLUB=y
> (same as native). Documentation/vm/slub.txt has some info on adding
> debugging stuff there, e.g. adding slub_debug to the command line. It
> doesn't look like rebuilding with the other two option would initially
> be useful (the first is equivalent to the command line option anyway)

I'm wary of enabling slub_debug by default; the document says it's okay to
enable individual items on runtime:

% ls -l /sys/kernel/slab/net_namespace/ | grep rw
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 min_partial
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 order
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 poison
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 reclaim_account
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 red_zone
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 remote_node_defrag_ratio
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 sanity_checks
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 shrink
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 store_user
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 trace
-rw-r--r-- 1 root root 4096 2013-02-20 15:41 validate

>From the documentation, I probably want:
        U               User tracking (free and alloc)

But which of the above files corresponds to that, 'store_user' or?
I'll have to go look at the source.

-- 
     2. That which causes joy or happiness.