Combating nginx 499 HTTP responses during flash traffic scenario

Tom Burns tom.burns at jadedpixel.com
Thu Nov 29 20:57:48 UTC 2012


On Thu, Nov 29, 2012 at 3:41 PM, Eric Wong <normalperson at yhbt.net> wrote:
> I fixed up some minor line-wrapping, signed-off, and added your
> quote above to the commit message.  Pushed as
> commit 5c700fc2cf398848ddcf71a2aa3f0f2a6563e87b
> to git://bogomips.org/unicorn.git
>
> I'll tag and push a 4.5.0.preview1 gem soon

Excellent news :)

On Thu, Nov 29, 2012 at 3:30 PM, Lawrence Pit <lawrence.pit at gmail.com> wrote:
> we haven't tested it with older browsers
> such as IE6 (we do not have that requirement) nor with all of the API
> clients we see (java, python, etc., we do have that requirement).

With nginx and proxy_pass you can buffer the entire request/response
from the rails application, so clients will not notice any difference.

>
> Hopefully we're bringing that patch into our production environment within
> the next couple of weeks, and then we should see results later that week.
>
>
> Tom Burns wrote:
>>
>> Unexpectedly the patch also helped us ward off a DoS attack where the
>> attackers were disconnecting immediately after making a request.
>
>
> Awesome!
>
> The only thing that worries me is that Eric calls this a "nasty/crazy
> solution". :p Not sure why. Seems to me it's a rather smart solution that
> perhaps in the future would deserve to be the default.
>
> A related question: is it possible to get insight in what's in the unicorn
> queue and for how long requests have been queued there?

We use New Relic for response queue monitoring.

https://newrelic.com/docs/features/tracking-front-end-time

Cheers,
Tom


More information about the mongrel-unicorn mailing list