[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: how far ha_proxy cartridge itself can scale, in openshift-origin??

+++ Prasad Varakur [06/03/14 15:17 -0800]:

I am trying to find scaling limits of an application. Looks like all
web-requests first land at web-server, then go to ha_proxy, from where
messages are distributed to multiple gears of my application.

In such design, if I have too many incoming web-requests, that can
choke my web-server, and/or haproxy. How to handle this?? How can I
configure multiple web-servers and multiple haproxy cartridges for my

Btw, I am using latest openshift origin, and looked at the user guide
and cartridge guides. Couldn't find any detailed documentation about
the webserver and haproxy cartridges.  Appreciate any help, or
reference to documentation.

I'm assuming we have community documentation for this too but the
OpenShift Enterprise docs on this can be found here:


The short of it is that you configure a broker plugin that then drop
messages in ActiveMQ.  From there you write a listener script that
reads the message and then contacts your load balancer of choice and
does what you want.

Near the bottom of that section it's explained how multiple HA proxies
can be enabled for an application.  What you'll likely want is for
your top level proxy to load balance and health check the multiple HA
proxy gears.


PS: see what we are doing.. Billions of animals are living in H E L L
every day waiting for the best day of their life, the slaughter day.
Be human, stop animal cruelty.

users mailing list
users lists openshift redhat com

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]