What is our primary use case?
I have been using Squid Proxy Cache Server for nine years in my current company. Before that, I used it overseas from early 2008 to 2010, so I have approximately ten years of experience overall.
In our infrastructure, the main use case for Squid Proxy Cache Server is serving server-side internet connectivity rather than user connectivity. When a server needs to access the internet, it uses Squid Proxy to do so, and we implement ACLs on the Squid server as well.
For example, we recently built a Microsoft server that needs to access the internet to patch, update, and perform other functions, and this is where Squid Proxy Cache Server really helps my team. We gather the information of the server source IP and the destination they want to access, such as abc.microsoft.com or xyz.microsoft.com. Then we create ACLs in the Squid configuration file with source and destination lists, attach them with the policy, and that is how we use it.
We have hundreds of servers in our environment, and I add to their requirements daily. Each server's requirement with Squid Proxy Cache Server is different, as they want to access different sites. We strictly control them, and it works both ways. It is both cache and internet control, mostly internet control.
In our environment, we use Squid Proxy Cache Server to control server-side access to the internet. On top of that, caching is also a benefit because the same cache is used by other servers. We download images from Microsoft, and if they are in the cache, the next download will be faster.
What is most valuable?
I have experience with Cyberoam as well, which also uses Squid as a caching server. Proxy servers have two components: one is caching, and another is internet access or URL database. The content or destination that has already been used can be saved in the cache. This is the best feature: cache in and cache out.
I used Squid Proxy Cache Server in my previous organization with delay pools to control bandwidth. We reviewed our MRTG or Cacti graphs before and after implementation, and after that implementation, we saw around twenty to twenty-five percent of internet traffic saved.
It is very difficult for us to determine if the twenty to twenty-five percent savings is from caching or from internet control. However, what we found is that before implementing Squid, we had around one hundred Mbps bandwidth utilization on average. I cannot remember the exact figures, but after that implementation, we found around seventy to eighty percent of internet utilization according to the graphs, while the user base and everything else remained the same. There was no degradation or employee shortage. Everything was as usual. Therefore, we feel that we saved around twenty percent bandwidth.
What needs improvement?
I did not use the AI capabilities of Squid Proxy Cache Server, but everything is moving toward AI. There should be some AI capabilities so that we can use delay pools or bandwidth quotas, and based on that, there should be auto-scaling. If the quota is filled, then additional chunks should be added to the quota automatically.
I would rate it ten if we had a dashboard or visible interface with Squid Proxy Cache Server. Because we are using third-party tools such as Cacti and MRTG to see utilization, and other third-party monitoring tools, there should be built-in tools in Squid so that we can measure performance and see results. If that were available, I would give it a ten.
What do I think about the stability of the solution?
I have not encountered any stability issues with Squid Proxy Cache Server so far. It is stable.
What do I think about the scalability of the solution?
In our environment, we implemented scaling differently with Squid Proxy Cache Server. We did not implement anything inside Squid for scaling. Instead, we placed a load balancer in front of Squid Proxy, and when anything requires scaling, we simply add a new Squid Proxy behind the F5 load balancer. That is how we handle scaling. However, I believe there should be automatic scaling with AI. I have not tried that.
How are customer service and support?
We never used the Webmin GUI for Squid Proxy Cache Server, but for a Linux expert, it is easy to control from the command line. I never have experience using the GUI. I have rarely used Webmin, but I did not discover how Webmin is useful. There should be a GUI interface or UI so that administrators can control it more easily.
I always thought Squid Proxy Cache Server is open source, so I never considered that there should be support. However, it would be great to have support similar to Red Hat Enterprise Linux. Before it was open source, but later it became enterprise. Currently, I do not know if there is a professional support team available for Squid Proxy Cache Server, but it would be helpful.
So far, we have not needed to reach out for help regarding support for Squid Proxy Cache Server.
Which solution did I use previously and why did I switch?
Before, we used Blue Coat Proxy, Cyberoam Proxy, Forcepoint, and other solutions. The reason we switched to Squid Proxy Cache Server is cost. Those professional proxy solutions have huge yearly cost involvement, and currently with Squid Proxy Cache Server, we are using it at no cost.
We evaluated a few solutions: Blue Coat, Cyberoam, and Squid, mostly these three. We found that Squid is the lowest-cost option. We also tried Forcepoint. We put Forcepoint, Blue Coat, Cyberoam, and Squid into a matrix and table, and based on all those factors, we found that Squid is good in terms of cost. We also conducted research and found that everybody uses Squid as their backend caching solution.
How was the initial setup?
Squid Proxy Cache Server is deployed on-premises in our organization. We have approximately six Squid Proxy Linux machines, and we have an F5 load balancer that controls those six Squid Proxies. High availability can also be implemented, but we have not tried that.
What was our ROI?
I do not have specific data regarding return on investment with Squid Proxy Cache Server. However, in general, I feel it is useful. Mostly, I do not see any cost involved except for the hardware cost.
What's my experience with pricing, setup cost, and licensing?
For our environment, we are using the free open source Squid Proxy Cache Server, so there is no licensing cost. Our only cost is for hardware, such as the Linux machine. In terms of costing, we do not feel much impact. It is very minimal.
What other advice do I have?
I advise others looking into using Squid Proxy Cache Server that if they are proficient in Linux and have a good Linux administrator, they should go for it.
Everything is good regarding Squid Proxy Cache Server. I would just advise that there should be a professional web interface for Squid so that we can see the consumption of the proxy. My review rating for this product is eight out of ten.
Which deployment model are you using for this solution?
On-premises