# Load balancing question / recommendation

**URL:** <https://www.ruby-forum.com/t/load-balancing-question-recommendation/200526>\
**Category:** NGINX\
**Created:** [December 14, 2010, 4:48pm UTC](https://www.ruby-forum.com/t/load-balancing-question-recommendation/200526 "2010-12-14T16:48:15Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Dobai-Pataky\_BSSSSl](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/d/77aa72/32.png) [@Dobai-Pataky\_BSSSSl](https://www.ruby-forum.com/u/Dobai-Pataky_BSSSSl)\
**Post date:** [December 14, 2010, 4:48pm UTC](https://www.ruby-forum.com/t/load-balancing-question-recommendation/200526/1 "2010-12-14T16:48:15Z")

</div>

Hi All,

Currently we have a single web server (nginx) with multiple php  
application  
servers (php-fpm) and the load is being handled well. The raid  
controller  
failed last week on the web server and we were down for a few hours.  
Adding  
a hardware or software based load balancer in front of the web server  
and  
using additional web servers, although still represents a single point  
of  
failure, will further increase the availability of our system should one  
of  
the web servers fail. We will be fully redundant in the future.

Our provider is recommending:  
[CAI Networks - WebMux Layer 2, 3, 4, 5, 7 Solid State Load Balancer and SSL Accelerator IPV4 and IPV6](http://www.redhillnetworks.com/products/webmux/load-balancer.htm) as a  
hardware solution, which I’m fine with, but I wanted to put it to this  
group, is it better to use a hardware solution such as this or to add  
another server with Nginx as a reverse proxy to a group of web servers?

> From a hardware point of view, we don’t have any choice in terms of a load  
> balancer (other than putting up another server with software on it).

Looking for real life experiences / recommendations and suggestions.

Thanks!

---

<div class="post-metadata">

**Author:** ![iberci](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/i/a3d4f5/32.png) [@iberci](https://www.ruby-forum.com/u/iberci)\
**Post date:** [December 14, 2010, 4:55pm UTC](https://www.ruby-forum.com/t/load-balancing-question-recommendation/200526/2 "2010-12-14T16:55:49Z")

</div>

2010/12/14 Ilan B. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid)

> Hi All,  
> Currently we have a single web server (nginx) with multiple php application  
> servers (php-fpm) and the load is being handled well. The raid controller failed  
> last week on the web server and we were down for a few hours. Adding a hardware or  
> software based load balancer in front of the web server and using additional web  
> servers, although still represents a single point of failure, will further  
> increase the availability of our system should one of the web servers fail. We  
> will be fully redundant in the future.  
> Our provider is  
> recommending:[High performance load balancer -- CAI Networks′ WebMux](http://www.redhillnetworks.com/products/webmux/load-balancer.htmas) a  
> hardware solution, which I’m fine with, but I wanted to put it to this group, is  
> it better to use a hardware solution such as this or to add another server with  
> Nginx as a reverse proxy to a group of web servers?  
> From a hardware point of view, we don’t have any choice in terms of a load  
> balancer (other than putting up another server with software on it).  
> Looking for real life experiences / recommendations and suggestions.

hi,

you can check on LVS with direct routing if all af your servers are on  
the same subnet. (without direct routing otherwise). It works like a  
charms, it has failover and it can handle a LOT of traffic. Delivering  
1gbps should be easy. If you have to handle a huge amount of traffic,  
you can loadbalance on several LVS instances with DNS round robin.

server1 (LVS1 master, LVS5 slave)  
server2 (LVS2 master, LVS1 slave)  
server3 (LVS3 master, LVS2 slave)  
server4 (LVS4 master, LVS3 slave)  
server5 (LVS5 master, LVS4 slave)

it rocks ! 🙂