Infrastructure Monitoring and Observability Platform

LogicMonitor  is primarily used for observability and data collection.

As a service provider, I believe the biggest benefit from LogicMonitor  for my clients is that it provides full-stack infrastructure observability. It is a very robust solution that combines all types of telemetry in terms of infrastructure observability. In addition to that, the latest additions to their portfolio in the last couple of years, with the acquisition of what is now called Edwin AI for event management and subsequent enhancement of that very same part of the portfolio to increase event correlation and AI-based features, represent a very big benefit. Last year, they acquired Catchpoint , which is also a tremendous benefit in terms of addition to end user and digital experience. They are building a very strong suite that is not very commercially attractive from a service provider point of view, but definitely, if the customer is looking for premium infrastructure observability, it is one of the most mature solutions that I have seen so far.

In terms of functionalities, I do not work much with Dynamic Service Insights for real-time visibility. I understand what they want to do. The mapping of infrastructure and applications and cloud services, and aligning the dependencies to show the business services that the customer has is absolutely fantastic if it is an end user customer that is a medium SOHO or medium segment customer that does not have a very big environment. Personas such as CIOs or SREs can see at a glance the service health of their business services. It is an acceptable functionality.

I do see negatives and areas for improvement in LogicMonitor. Obviously, there is always space for improvement. First of all, I believe the commercial aspects require improvement. At the moment, if I subscribe to everything on the platform, it becomes quite a large cost. Although it has many benefits, an observability solution, even a full-stack observability solution, should not exceed a maximum of eight percent of the annual budget of the environment. If I buy everything, I will definitely exceed that percentage.

In addition to that, I believe one of the biggest drawbacks that does not make LogicMonitor stand out like Dynatrace  or DataDog is that they have not focused on the APM  side. Application Performance Monitoring  is crucial; if we have a customer with many applications needing instrumentation, or even SaaS platforms monitoring, this is where LogicMonitor loses ground. The APM part is somewhat lacking. They have made some attempts, but it is very rudimentary and I would not even categorize it as a solution to be honest. I would very much like to see them invest in this area to start competing against DataDog.

The principal thing that differs LogicMonitor from Dynatrace  and DataDog is the APM side. Application instrumentation is essential, as only by adding that will you become a full-stack observability solution. At the moment, they are only an infrastructure full-stack and infrastructure-focused solution.

I started using LogicMonitor eight years ago.

LogicMonitor is ninety-nine percent stable all the time.

LogicMonitor is highly scalable. I would describe it as very scalable.

In terms of customer service, I would divide my answer into two points. Professionally, I think they have very knowledgeable staff in their environment, and support is good most of the time. To be specific, I have no concerns; I think they are doing a good job. Occasionally, we face some challenges, but overall, I see it as very good support. I have experienced far worse.

In terms of deployment procedure, I find it very straightforward. The main instance is up and running in twenty-four hours. Deploying collectors is also substantially easy, so I do not foresee much problem there; that is actually an accelerator.

The effectiveness of Edwin AI can be measured depending on what I am trying to achieve. There are several aspects of Edwin AI. One is event management, which is simple event management as they started, and it is quite robust. The measurement of the effectiveness is the reduction of the number of incidents that are created, so it involves basic root cause analysis in this case. Then I have the second part where the AI capabilities, or in other words, AI intelligence comes in, and they position their own LLMs to give an AI flavor of the analysis of the issue. In this case, they enhance the correlation based on ingestion of their own service graphs, either taking from the ServiceNow  topology maps or taking from the additional information from their own topology that they have in LogicMonitor, which they now call Envision , but it is LogicMonitor Core. They provide quite a substantial enhanced capability in terms of correlation. Subsequently, the generative AI algorithm that they have is useful to summarize the issue and provide outcomes. With the addition of the chatbot capability, it is quite robust and a good functionality. The last part is basically the ability to have, which they have launched very recently, the AI Control Tower , which gives me the ability either to use the AI agents that are already in place, to substantiate the agent AI point of view, or to create my own, which is a very good addition.

The only drawback is that I am limited to use the LLM from LogicMonitor. While the industry nowadays is more oriented for the user to choose which LLM they want to use, that is a current limitation. This is a capability in development. I know they will launch this because the market is demanding and their competition is already doing that. But I do not see a showstopper for the adoption. I do believe that they should launch this capability and enable us to use our own LLMs. For instance, if I have something like either Copilot from Microsoft or Bedrock from Amazon or Gemini  from Google, or even my own hosted, self-hosted option, I want to use that. Why would I need to pay for additional LLM tokens if I already have my own LLMs? This is a very concise standardization of the industry at this point in time. Many other players are already doing that and this will be no different.

Incident management is part of the Edwin AI point of view. All three functionalities that I have just described are part of incident management. Basically, what it is, is I create an incident in whatever ITSM  tool or solution that I am using, based on the correlation on the output of Edwin AI. I am referring to autonomous learning. Autonomous learning is basically the feature of Edwin AI. They have their own LLM which keeps learning the patterns, which I consider more a machine learning algorithm than actually AI. But it is sufficient at the moment; I think they are on a good path since they launched this capability about six or seven months ago. I would say that I have not seen many hallucinations, at least in the environments that we have deployed, which means I give a positive note on that.

Regarding LogicMonitor's impact on our mean time to resolve incidents (MTTR), I want to emphasize that where we have deployed Edwin AI, we see our MTTR decreasing substantially. I am talking about real numbers and not the numbers they provide. The statistics will circulate depending on the maturity of the environment, since that is a KPI that must be considered. The data foundation is the principle of the effectiveness of MTTR reduction with Edwin AI. In a normalized mature environment, I would say that I see a reduction in resolution times of forty-two to fifty percent. This happens because we substantially reduce the number of incidents that are generated. Therefore, there is an obvious reduction, but the industry standard is slightly higher or at least demands to be a bit higher. Again, it comes back to the point that I just mentioned; it depends on the maturity of the customer.

Regarding whether LogicMonitor's solution is rather expensive, I say that if I evaluate the value, it is substantial. First of all, you need to understand that I am speaking from a service provider point of view, not that of the end customer. Even for the end customer, it would be slightly difficult if they buy everything. However, the value drawn from it is considerable. If they have the budget to spare, it is a very good solution. From an MSP point of view, it becomes unattractive because it inflates our commercial proposals to the customer. The customer, in a managed services context, does not see the solution bringing cost XYZ; they see the overall cost of the service we are providing. Not to mention that nowadays, there is a lot of competition, especially with the trend towards open source, where customers sometimes choose to go open source instead of buying an enterprise product, even one as mature as LogicMonitor.

We do not purchase anything from LogicMonitor through the AWS Marketplace ; we buy licenses directly. Our labs, however, do run on AWS , so as an end-user consumer of AWS , we have many services there. There are some challenges in gaining complete visibility across my hybrid infrastructure and I observe that the Envision  platform addresses these challenges. Usually, I do not see many challenges onboarding the environment into the platform because it can adapt quite easily. If it is not there, we can always ask LogicMonitor to address these points, which they frequently do in the majority of cases, either in a quarter. So I do not encounter problems with that. My main concern nowadays is how to extract data from LogicMonitor. I need to combine data in a unified manner in the era of data lakes to mix operational and non-operational data outside them. However, extracting data is particularly difficult due to their low API rate limits. I understand the reason behind the low limits, ensuring no one overloads the platform, but it makes integration stressful. I would rate this review an eight overall.

My use case for LogicMonitor  is to create alerts, specifically to create PagerDuty alerts and Slack alerts for high usage of memory, high performance, and similar issues, ensuring the infrastructure is up to date.

I can say it has reduced my mean time to resolve by almost 30-35%.

The best features of LogicMonitor  are that the monitoring is excellent compared to other products in the market. The alerts are accurate, and it is agentless; you can automate it, and we don't need human intervention, as it automatically discovers the devices so you don't have to do it manually. It supports a hybrid cloud model and multi-cloud model monitoring.

I have used the Dynamic Service Insight feature for real-time visibility for one of my clients to monitor individual devices' health. They had a lot of devices, but they wanted to monitor them individually and automatically group the devices to check their alerts, check their service level score, and check the dashboards in real-time.

I have used LogicMonitor's Edwin AI for diagnosing root causes and remediation to reduce the noise alerts that come in. It reduced the noise alerts significantly, almost close to 90%.

Areas that have room for improvement in LogicMonitor include that the product should be a plug-and-play system. The dashboards should be customizable, as there are products in the market such as Grafana  which are quite customizable. I think the product is expensive when looking at products in the same segment in the market. The reporting should be customizable and more advanced when looking at Edwin AI. Log management should be easier, similar to Splunk or Elastic.

I have been using LogicMonitor for around two years.

Regarding stability, I can say there is mostly no downtime, so I rate it a 10; I think 10 would be a good number.

The solution does require maintenance, such as checking the devices, discovery of the devices, and checking if any ports are blocking, but usually it is agentless.

I would rate the technical support of LogicMonitor as eight.

I estimate that the return on investment for me and my clients is around 90%.

I compare LogicMonitor mostly with Grafana  and Splunk. Splunk has strong analytics capabilities, Grafana is more of a cloud infrastructure application, and Nagios is legacy software in the market for network monitoring. Since LogicMonitor performs much the same job, I expect it to do the core job better.

I would suggest DataDog, but if the budget is good, DataDog would be a good option or Splunk. However, looking at the organization level and budget, LogicMonitor is also a value for money product.

Regarding the autonomous learning feature, I have not used it much.

The challenges I face in gaining complete visibility across my hybrid infrastructure involve discovering the devices, as it's somewhat challenging and sometimes it misses a couple of devices.

My clients include small, medium, and enterprise businesses, as I work with all three of those types.

In my organization currently, I can say the entire team, which is almost 40-45 members, works with LogicMonitor.

I am providing this review with an overall rating of 9.