We started using SentinelOne Ranger, but we found two problems. Perhaps they are particularities, but they should be addressed as they may change the minds of other companies that are considering this feature.
The first problem is that, while it scans all the assets that are on the network, when it comes to discerning whether an asset is a server or a laptop, it tends to fail. It does not have a very high level of precision. We have experienced problems when reporting these types of assets to those responsible for installing the agent, and then they tell us, "Hey, this is not a server, this is a fax," or "this is a printer." When things like that happen, we lose credibility.
The other issue that we saw with the functionality of Ranger is that if, for whatever reason, you have a product with SentinelOne installed but it is on a client's network, the SentinelOne agent starts scanning the ports and the network and goes to a honeypot. As a result, the client may think that it is being attacked because someone has reached its honeypot, when it’s actually us on the client's network. When you don't know that this is happening, it can generate conflict and tension with the clients. Once you know about the problem, you can deactivate that process, but sometimes it can have a negative impact.
Ranger does provide me with visibility of the network, but not completely because the assets it scans are often mistakenly identified regarding what type of device they are. A SentinelOne agent is worth a lot of money, and there is no point in putting it onto a printer, for example. It should have the ability to go a little further and be more precise.
Another very clear area for improvement, one that I don't understand why they haven't deployed it yet, is a self-updating SentinelOne agent. The agent has a version, and what SentinelOne proposed up until one year ago is that you had to be proactive in consulting the dashboard to see if your agent had reached end-of-life and then update it. Now, they've released a new feature where I believe you can schedule updates, so it makes perfect sense for the agent to update itself without any action on our part, and never go out of version. By simply connecting to the network it should be able to download and update.
This idea is not critical because SentinelOne updates many versions of the agent and, when one becomes obsolete, it does not mean that it no longer works. But this is something that SentinelOne should know how to work with. A solution could be that if you do not have the ability to auto-update the agent, SentinelOne would directly tell you which agents are not updated. That way, we would not have to go to the documentation, look at the dashboard, and filter the agents by version. It would be great if it were able to tell if the operating systems are unsupported so that we wouldn't have to look in the official documentation at whether the Windows Server is outdated or not.
If the agents self-updated, maintenance due to the update process would be minimal.