We have a couple of servers on-premises to gather the logs from our devices. We have a lot of devices including vendor-agnostic collectors that will, for example, collect syslogs from our Linux host. The logs are then sent to the Devo Relay, which encrypts the data and sends it to the Devo Cloud.
What we send to Devo includes all of our Unix-based logs. These are the host logs, as well as logs from a lot of the network devices such as Cisco switches. Currently, we are working with Devo to set up a new agent infrastructure, and the agents will collect Windows event logs.
We were using a beta product that Devo provided for us, which was based on an open-source platform called Osquery. That did not quite work for the volume of logs that we have. It didn't seem to be able to keep up with a large number of servers, or the large amount of Windows event log volume that we have in our environment. We're currently working with them to transition to an Xlog and use their agents, which work really well to forward the logs to Devo.
We also send cloud logs to Devo, and they have their own collector that handles a lot of that. It basically pulls the logs out of our cloud environment. We are sending Office365 management logs, as well as a lot of Azure PaaS service logs. We're sending those through an event hub to Devo. We are currently working on onboarding some AWS logs as well.
We have several corporate locations, with the main location in the US. That is where the majority of our resources are, but we do also have Devo relays stood up in Canadian, Australia, and India. These locations operate in a way that is similar to what is described above, although on a smaller scale. They're sending all of their Unix devices and syslogs to the relay, and then I believe only Australia at the moment is using agents to pull from Windows logs. Canada is using a different SIEM at the moment, although that contract is about to expire, so then we'll onboard their Windows event logs as well. India does not have any Windows servers that need to have an agent for collecting logs, so just send the Linux and Unix logs over the relay to Devo.
Our main use case and customer base are our security operations center analysts. A lot of our process was built up and carried over from our previous SIEM, LogRhythm. We have an alerting structure built out that initiates a standard analyst workflow.
It starts when you get an alert. You drill down in the logs and investigate to see if it's a false positive or not.
We are in the process of onboarding our internal networking team into Devo, and we are gathering a lot of network logs. This means that they can monitor the health of our networking infrastructure, and at some point, maybe set up health alerts for whatever they are looking for.
We have another team that is using Devo, which is our internal fraud team. They're very similar to stock analysts, where they just look for suspicious events. They are especially interested in tax filing and e-filing. We gather logs for that, and they go through a really deep investigative workflow.
One of the immediate improvements that come to mind is the amount of hot, searchable data. In the SIEM we had before, we were only able to search back 90 days of hot, searchable data, whereas here we have 400 days worth. That definitely has improved our threat hunting capabilities.
We're also able to ingest quite a bit more data than we were before. We're able to ingest a lot of our net flow data, which if we had sent that to our previous SIEM would have brought it to its knees. So the amount of data that the analysts are able to see and investigate has been a really big beneficial use case. I'd say that's the biggest benefit that it's provided.
I myself do not leverage the fact that Devo keeps 400 days of hot data to look at historical patterns or analyze trends. A lot of times I will look at that to see the log volumes, the traffic, make sure there are no bottlenecks as far as how log sources are sending to Devo. I would say that the analysts definitely for certain cases will go back and try to retroactively view where a user was logging in, for example. At the moment, we haven't really had a use case to push the limit of that 400 days so to speak, and really go really far back. We definitely use the past couple of months of data for a lot of the analyst cases.
This is an important feature for our company especially with the recent SolarWinds attack, which was a big deal. We did not have Devo available, but because that happened so far in the past, it was a struggle to pull that data for it to look for those IOCs. That was definitely a really big selling point for this platform with our company.
Devo definitely provides us with more clarity when it comes to network endpoint or cloud visibility. We're able to onboard a lot of our net flow logs. We are able to drill down on what the network traffic looks like in our environment. For the cloud visibility, we're still working on trying to conceptualize that data and really get a grasp around it to make sure that we understand what those logs mean and what resources they're looking at. Also, there's a company push to make sure that everything in the cloud is actually logging to Devo. As far as cloud visibility, we as a company need to analyze it and conceptualize it a little bit more. For network visibility, I would say that Devo's definitely helped with that.
The fact that Devo stores the data raw and doesn't perform any transformation on it really gives us confidence when we know that what we are looking at is accurate. It hasn't been transformed in any way. I'd definitely say that the ability to send a bunch of data to Devo without worrying about if the infrastructure can handle it definitely allows us to have a bigger and better view of our environment, so when we make decisions, we can really address all the different tendencies. We're collecting a lot more types of log sources than we were before. So we can really see all sides of the issue; the vast amount of data and the ability to really take our decision and back it up with the data, and not just random data but we can use a query and display the data in a way that backs up the decision that we're making.
Devo helps to release the full potential of all our data. The Activeboards like the interactive dashboards that Devo provides really help us to filter our data, to have a workflow. There are a lot of different widgets that are available for us to visualize the data in different ways. The Activeboards can be a little slow at times, a little bit difficult to load, and a little bit heavy on the browser. So sometimes the speed of that visualization is not quite as fast as I would like but it's balanced by the vast amount of options that we have.
That's one of the big things that like all security companies, security departments really purported having that single pane of glass. The Devo Activeboards really allow us to have that single pane of glass. That part is really important to us as a company to be able to really visualize the data. I haven't found the loading speeds have become a significant roadblock for any of our workflows or anything, it's an enhancement and a nice to have.
We all want everything faster, so it's definitely not a roadblock but the ability to represent the data in that visualized format is very important to us. It's been really helpful, especially because we have a couple of IT managers, non-technical people that I am onboarding into the platform because they just want to see an overall high-level view, like how many users are added to a specific group, or how many users have logged in X amount of days. The ability to provide them not only with that high-level view, but allow them to drill down and be interactive with it has really been super helpful for us as a company.
Devo has definitely saved us time. The SIEM that we were on before was completely on-prem, so there were a lot of admin activities that I would have to do as an engineer that would take away from my time of contextualizing the data, parsing out the data, or fulfilling analysts requests and making enhancements. The fact that it is a stock platform has saved me a ton of time, taking away all those SIF admin activities.
I wouldn't say that it really increased the speed of investigations, but it definitely didn't slow it down either. They can do a lot more analysis on their own, so that really takes away from the time that it takes to reach out to other people. If you went back 90 days, you had to go through a time-consuming process of restoring some archives. The analysts don't have to do that anymore, so that also cuts off several days' worth of waiting. We had to wait for that archive restoration process to complete. Now it's just you pull it back and it's searchable. It's right there. Overall, I would say Devo has definitely saved us a lot of time. For the engineering space, I would say it saves on average about one business day worth of time every two weeks because a lot of times with on-prem infrastructure, there would be some instances where it would go down where I'd have to stay up half the night, the whole night to get it back up. I haven't had to do that with the Devo platform because I'm not managing that infrastructure.