> 
> Which one would be good for capacity\load? I have a very, very large
> environment. I have 220,000 users on 8 Gig to the INTERNET. I am
> running a load balancer, ipvsadm (Direct Routing) with 20 proxies
> behind it. I am interested in handling load.
> 
> Michael
> 
> 

I'm working in a close environment + web filtering (e2guardian) - less 
bandwidth 
Before beginning be aware of SMP limitations, 
http://wiki.squid-cache.org/Features/SmpScale (look at "What can workers 
share?" )

I tried multi-CPU, to be honest not very recently, but the limitations were 
annoying for me and I encountered some issues with high load (crashes)

For the moment I can still manage with hardware over 5 years old, squid core 
reach 50 % of cpu usage with 700/900 request/s, diskd, basic + digest 
authentication, delay pool and caches 2 x 300 Go. The web filtering are using 
the other cores. 
The CPU usage was really reduced but without SMP I do not have any problem at 
all, I known no pain no gain but ...

My advice, more than CPU buy RAM to support thousand of users you will never 
have enough RAM ! (64 Go if you can) and really fast disks - dedicated to cache 
-

Also there are some little things that you can do to help you to improve your 
capacity, a squid restart each day, tuning the network, tuning the system, etc 

The bandwidth is a point, but the requests by second is THE point 

20 users can get some 4K videos on youtube your bandwidth is consumed but your 
load average still close to 0.0
20 users with a spyware (or something similar) can push hundreds of requests by 
seconds, in this case your bandwidth can still stable - with the cache - but 
your load average increases

It depends   

Fred    


 
_______________________________________________
squid-users mailing list
squid-users@lists.squid-cache.org
http://lists.squid-cache.org/listinfo/squid-users

Reply via email to