Varnish crashing when system starts to swap

Eden Li eden at
Tue Apr 29 09:47:36 CEST 2008

On Mon, Apr 21, 2008 at 3:08 PM, Poul-Henning Kamp <phk at> wrote:
> In message <BCA2D4755E06EF44A0278A1D9B217E3F08171B3361 at exchange01.ex.local>, Ca
> lle Korjus writes:
>  >Varnish looks fine until it's had abour 1,5 million requests, then
>  >we can see the kswapd0 and kswapd1 start working and load average
>  >rises to about 200 and the machine gets totally unresponsive. Top
>  >shows a lot of cpu beeing spent on i/o waits and varnish child
>  >process restarts sometimes. In best
>  >case the process restarts and the server starts behaving within 5
>  >minutes but sometimes varnish dies completely. One thing we have
>  >noticed is that the reserved memory for varnish keeps rising and
>  >when it crashes it is usually around 14G.
>  You didn't say which varnish version you are using, but we have fixed
>  a number of memory leaks in -trunk recently.

I've just run into this exact issue varnish 1.1.2, except that varnish
was running for over a month until kswap started acting up.  Also, our
backends are not on the same machine.

I tried running different revisions of varnish from trunk on our
machines (r2634 and r2436 on Linux CentOS 5.0), but it keeps asserting
when daemonizing the process (varnishd.c:557).  Using an export of the
unstable branch kept at seems to be

Do you know if this branch has some of the memory leak fixes you mention here?

More information about the varnish-misc mailing list