Varnish 3.x Expiry Thread overload

Lasse Karstensen lkarsten at varnish-software.com
Tue Aug 27 11:28:22 CEST 2013


On Mon, Aug 26, 2013 at 04:27:34PM +0200, Ricardo Bartolome Mendez wrote:
> There are no bans in use. I think we can manage to stop sending small
> objects to a subset of servers and see how it behaves, even unfortunately
> we'll need to wait few weeks until we check if it helps or not. We'll also
> give a try to -smalloc to see how it behaves.

Great. Please report back if the malloc backend works as expected.


> Shall I understand that you are actively looking into the issue because a
> customer of yours is being affected or shall I open a ticket in Varnish
> Trac? Is it a known issue? Is there already an existing ticket? I'm willing
> to help, so don't hesitate to ask me any kind of information that can help
> you to spot the issue.

The customer got things working with malloc and was happy afterwards.

If our gut feelign is right, this is a situation where the assumptions about
the working set made during implementation is not entirely according to how it
is being used.

I don't believe there is a trac ticket for it. Tracking down these things are
pretty resource intensive. Create one if you have time to follow through on
in, but I'd expect that the 4.0 release gets priority the next few months.

If you need this fixed in a shorter time scale, there is commercial support
available. Ping me off-list if this is of interest.


-- 
With regards,
Lasse Karstensen
Varnish Software AS



More information about the varnish-misc mailing list