What is our primary use case?
Cortex XDR by Palo Alto Networks is more than an antivirus. It collects data and sends it to the Cortex Data Lake, which serves as your central location where you can manage logs and investigate incidents. If something is being blocked, you can investigate further to understand what is blocking on the second or third layer of the application. For example, if you have a trusted application that is not working as it should, Cortex XDR can help identify the issue.
Cortex XDR not only verifies the executable that you are running, but it also analyzes all the dependencies that this executable is trying to call and sends all this information to the cloud database, to Cortex Data Lake. The product inspects the program that is running and also its dependencies. It also inspects the connections where that executable is connecting and from where it is connecting, along with all its dependencies.
After that, it is more than an antivirus, and what I can add is it depends also on the license that you have. If you have the basic license, there are some things that you cannot control. For example, you have SAP running, and SAP is calling for a DLL to print something on a specific printer. That DLL is not signed because it is a makeshift drive, a specific driver for that printer. When you try to print and it is not printing, Cortex XDR blocked it. If you have a basic license, the only thing that you can do is allow or not allow that dependency to run. If you have the Pro license, you can tweak further. You can create rules by behavior, not just allow everything from that application, but you can allow a specific behavior, just that behavior, that communication, that port. It will allow the communication to be made. If that DLL tries to communicate from a different behavior, from a different port, or if that DLL makes a different behavior, Cortex XDR will detect and block it. This depends on the license.
Cortex XDR is internet dependent, but you can make it offline. You have to install a middleware server that you have access to in Palo Alto to install, which is a broker VM. The broker VM has to have internet connections and will update Cortex XDR and retain the logs from all Cortex agents on the network and send it to the Cortex Data Lake.
In Cortex Data Lake, there is machine learning that is not in the agent per se, but in the cloud on your own tenant. When you buy a license, you have a tenant and Palo Alto gives you storage space for log retention. All these logs and behaviors that the agents on the endpoints collect are sent to Cortex Data Lake, which is on the cloud. Those logs are being read by machine learning from Palo Alto to determine those behaviors and specific behaviors from applications or even the system, and it will help you on the control panel from Cortex XDR. You will see all the incidents and be able to see incidents. All that data is going to your tenant, to your cloud, and yours only. Cortex Data Lake is a Palo Alto data center, but your data is your data. Your logs are your logs. They are not shared with anyone.
On the Data Lake, there is machine learning that inspects all these logs and therefore can detect specific behaviors that are happening on the endpoint. From the start, you can inspect specific behaviors, abnormal behaviors, or even incidents that may be a false positive or true positive. If there is malware, spyware, or something similar, it detects the behavior of the applications and dependencies from the second layer or third layer. If something is fishy, basically, it will be flagged. The cyber engineer has to check that incident and basically, give it a go or not, or block it, or create a new rule or deploy on all networks if it is a normal behavior from the application. For example, getting back to that example from the DLL of the printer or driver from a specific printer from SAP, and that DLL is trying to communicate to the printer on a specific port, Cortex XDR will detect it and it might, if it doesn't know what it is, block it. Then you can deploy and create a rule for that specific behavior and allow it on the network or part of the network or just an endpoint or two or three, depending on the group and depending on the scale that you have.
What is most valuable?
Cortex XDR by Palo Alto Networks has more than just antivirus capabilities. It verifies the executable and analyzes all its dependencies, sending this information to the cloud database, specifically Cortex Data Lake. It inspects the programs, dependencies, and connections, making it useful for blocking unwanted behaviors based on licenses, whether basic or Pro. Additionally, there is machine learning within Cortex Data Lake to detect specific behaviors from a cloud standpoint, which aids in producing accurate incident reports and managing them effectively on the control panel.
What needs improvement?
The main issue I could point out is the offline agents and the way that it is missing. Even if you create and isolate your network, for example, if you are proxying all your communications and not allowing any endpoint to connect directly through the internet, servers can have servers that are isolated, which is absolutely normal, yet the agent needs time to connect to the internet. It shouldn't be that way because it is a problem. You have to expose the servers to connect and then allow the upgrade or update that the agent may need or just to do the heartbeat and then close that network again. It is an unnecessary hassle because if you have a broker VM as a middleware, it shouldn't happen. The broker VM should do all the work.
For how long have I used the solution?
I have been using Cortex XDR by Palo Alto Networks for almost 10 years.
What do I think about the stability of the solution?
If you need more capacity, buy more licenses and just install the agent. It is as easy as that.
What do I think about the scalability of the solution?
I have not experienced any scalability issues.
How are customer service and support?
I have occasionally contacted technical support for Cortex XDR by Palo Alto Networks, and they are good. They are as fast as they can be. It depends on the type of contract that you bought and the type of license and support when you make your purchase. In the worst case, we are talking about maybe a maximum of four days. If you have a really high support level with a really good contract, you can expect a response in four hours. It depends on the type of support that you buy.
How would you rate customer service and support?
Which solution did I use previously and why did I switch?
I generally believe that Cortex XDR by Palo Alto Networks is probably the best in the market. I know Microsoft Defender, which is not good quality. I know IBM Reactor, which they tried to copy Cortex with, and they are doing it, but it is still not as good. It is better than Microsoft, yet too rudimental. I generally believe that Cortex XDR by Palo Alto Networks is probably the best in the market right now.
How was the initial setup?
Cortex XDR by Palo Alto Networks is truly easy to set up. I just give the agent depending on what type of system and what type of operating system you have to install. Either way, it is really easy. You can deploy it using a free tool. Instead of using Intune from Microsoft to make the deployment or the SCCM, if you do not have the license from Microsoft for those tools, you have a free tool that does it perfectly. That is the tool that I use and it is really easy to use, fast, and simple. On the user side, it is really easy. On the admin side, it can be overwhelming, especially on the first deployment. You can deploy the agents really fast. For example, for 600 endpoints, including desktops, laptops, and servers, you can deploy them in one day.
The main hassle is the creation of the profiles. You have to create profiles, endpoint profiles, policies, and group policies. Before that creation, you have to be certain what type of inventory you have. Before creating your policies, be sure what you want. If you want it by servers or if you want to do it by department, if you want to do it by department and servers and endpoints, create groups that are going to be easier to identify and manage. For example, for a few endpoints for marketing and commercial, those people do not work with IT tools. Yet, they go to Facebook, LinkedIn, Instagram, and whatever else, but they do not use IT tools. With that in mind, you create a specific profile for them. For IT people, it is the other way around. They have the right to use certain tools, but they do not need Facebook or Instagram.
That is important in terms of profiles. It will be easier to manage, and if something happens in the network, it will be easier to isolate and faster. This way, you have the group identified and by different policies, which I recommend. IT always has a different policy. Servers have one policy, and production servers and prediction servers have different policies from other servers. If you have different types of servers depending on the type of servers you have, Linux, Windows, if they are production servers, if they are database servers, if they are web servers, try to make groups and separate them. Create a group to differentiate them. It will be easier to manage. Instead of having all servers in one group and a few servers communicate differently than the others, it would flag something. It may block, which is something you do not want, or it may pass, which is also something you do not want either. If you have database servers, create a group for database servers. If you have web servers, create a group for web servers only. For domain servers, separate them. Give them a specific profile, the same as the desktops or laptops. Laptops for marketing and commercial have a specific profile for commercial laptops. For the administration, big boss, and VIPs, give them a specific profile or two profiles.
Do your work before creating those profiles and before starting to create those policies and profiles. If you have your inventory done and if you have your inventory in order, then you are good to go and can do it really fast. In two or three days, you have Cortex XDR done on all your network. The hardest thing is actually the administrative part, which is knowing your inventory, knowing your endpoints, and knowing your servers.
Which other solutions did I evaluate?
To tell the truth, I am not on that side. If a customer asks me, depending on the budget that they have, I do not know prices and I do not care about prices. If they ask me what I want them to put, I say Cortex XDR by Palo Alto Networks. If you do not have the budget, go to IBM Reactor. On third position, maybe Microsoft solution. But it is still in development. Reactor is not a final product and Microsoft Defender is worse. It is a hassle to work with. If you want an XDR on your network, put Cortex XDR by Palo Alto Networks. Any EDR or XDR or whatever, put Cortex XDR by Palo Alto Networks. If you can afford it, put it. If you can go further, buy a Pro license. I endorse buying the Pro license, specifically if you have more than 300 endpoints to control. It is way better. It is more refined and you have a lot more control, not only on the deployment, but also on the defense side.
What other advice do I have?
I believe Cortex XDR by Palo Alto Networks is probably the best right now in the market. If there are none better than Cortex XDR by Palo Alto Networks, then it deserves a top rating. I would not be so complimentary if it was not deserving of such praise. They have done a really good job. I gave this review a rating of 9 out of 10.
Disclosure: My company has a business relationship with this vendor other than being a customer. Partner