One thing I'd like to see in any APM, especially New Relic, is the ability to use distributed transactions. When one microservice calls another, it calls another database and microservice. The entire data visualization layer will not be able to correlate from one microservice from end to end and return on that path. Distributed transactions would be a great addition that would make life simpler. Unfortunately, no APM has that end-to-end capability. When I say "distributed transactions," I'm not only talking about the database level. It needs more and better visualization of communications across various microservices and integration with logs.
The UX/UI design of New Relic APM could be improved. The solution currently has some slow pages in terms of loading and viewing the pages, for example, the reports. The reports and other pages take a long time to load, so if that area could be improved, especially when looking for data, it would enhance New Relic APM.
Director of Performance Testing at a tech services company with 201-500 employees
Real User
Nov 29, 2022
Real-user monitoring would be helpful as it would help me to really understand the client-side performance of the application. Maybe for whatever reason, we have not got to explore a similar kind of feature in New Relic.
Director at a tech vendor with 1,001-5,000 employees
Real User
Nov 2, 2022
I would like the ability to set up certain dummy accounts and do the actual things that the customer is doing, without impacting the production environment. Only the read APIs are called from New Relic, not the write APIs. If we had a test account to do the write part of it, it would give us better monitoring. For example, if we are selecting the data for an existing account, we can do that part of the monitoring with New Relic. When we see failures and slowness, I would like there to be an option to do a deep dive into a collection of metrics to show the bottlenecks. It would be helpful if it didn't just state the problem, but indicate the areas to look at for a deeper resolution of the problem.
Documentation could be improved in New Relic APM, so users would have more clarity on configuring the dashboard. If New Relic gave better guidelines, users would find it easier to understand the metrics and features of New Relic APM. Another area for improvement is integration with Kubernetes. Currently, the process isn't user-friendly. It's challenging and lacks documentation for users to understand how to integrate New Relic APM with Kubernetes quickly. With multiple levels of Kubernetes dockers and other DBs on different clouds, it's tricky to gather all into New Relic APM on a single dashboard. What I'd like to see in the next version of New Relic APM is a single dashboard where you can easily view which applications fall under specific APMs. If there's a search feature where you can type in a keyword to find out if an APM is related to a particular application, that would be great.
IT operation manager at a tech services company with 201-500 employees
Real User
Aug 10, 2022
New Relic APM could improve error debugging and the correlation with the logs. We are receiving some alerts or alarms but we need to correlate with the error log, but it is difficult if it is more than seven months retention period, it is hard to trace. We need this especially for getting historical information. Within short periods of monitoring, we do not have a problem it is only when it crosses the retention period that we are not able to correlate the errors with the log.
We would like a dashboard feature to be created for this product. This would allow us to monitor both the front and back-end of our UIs performance, and then report on it. We would also like the solution to increase the details that are provided when there are system abnormalities.
It offers transactions, but it does not offer an endpoint-level insight at the URL level. When we get a request, we want to know what the life cycle of that service is, and where the cycle is. This is what I am trying to locate with most of the solutions now. I am trying to research how to find a cycle per endpoint and not at the service level. It is very difficult to award the service level cycles at an endpoint level. It is important for us to get new insights to create better hygiene around the business use cases. At the endpoint level, the visibility is not that great, and metrics are not available. It gives you a full view of the entire function's execution and not from the context of the URL altogether. Also the response time, the latency contribution, and the throughput contribution are areas that need improvement. You can get the throughput contribution from New Relic, but not the latency contributions. You cannot get it at all. These are the major limitations. When working with AppDynamics, I did not find any limitations, but the same can not be said with New Relic. The way that it classifies the actual services is a bit ambiguous. It's not perfect. For example, I see there are certain solutions that are listed as extra services, as a dependency, and still I find that among load contribution, it tries to show that those services separately, which is confusing. With the transactions, when it tries to show a type of "bufferHandler" from inside, it doesn't show what the nature of the request is. Especially with Microservices, it doesn't show what kind of method is present, which makes finding data very difficult. Instead, you need to go to the raw data. I think that defeats the purpose of using this tool. The transactions do not show the time consumed by the request, from the metrics execution perspective. It was suggested that I did not know how to read it but I have done all that I could. It is very difficult to relate to and requires a lot of experience and time to read through, which it should not. It should not be difficult to find the latency and throughput for the entire system when requested. It should not be difficult to develop the data that relates to the various types of execution. It should have complete exposure around the endpoints. The services-to-service dependency is fine but most of the startups have only one or two services that are all cycled. It does not provide you with a lot of help when you are showing that the two services are dependent. What all of the dependent endpoints are and how are the cycles being formed is information that should be available in most tools, but not with New Relic and some other tools.
I haven't yet had too much of a chance to read up on the product or to compare it to other solutions to see if there are items that are standard elsewhere but lacking in this product. I need more time with the solution to really talk about any shortcomings if they exist at all. It would be nice if there were pre-made dashboards. There should be out-of-the-box options that you can have. It would be nice to have something already done, so that you don't need to reinvent the wheel again, to create your own dashboard. The pricing might be high if you are a company that goes through a lot of data.
There has been some problem with the agent, and it is just not working well. It is not able to record information with the application server. They have been able to fix the issue, but it took quite a long time. This is the main issue in the APM products and also in New Relic. The mobile application monitoring has been pretty difficult to set up and also quite expensive. It should be a little bit easier and cheaper. Because it is pretty difficult and expensive, many customers don't take it.
Senior DevOps Engineer Individual Contributor at EML Payments Ltd
Real User
Jan 30, 2021
There really is nothing that stands out with New Relic. With the insight, I think it will be found lacking for its report aggregation capabilities. How granular I could go down at looking at certain data, especially related to the operations, is limited. The API integrations that they have for us to automate our configuration was fine, but I think for some of these tools, it was over-engineering for us to try and automate any of that. So, we just use the user interfaces.
The older view is much better than the new view that they have. We'd like to go back to that previous version. The user interface just isn't as nice as it used to be. The solution does not seem to be lacking any features. I can't say that I recommend any specific features for the next release.
Technology Competency and Solution Head at LearningMate
Real User
Nov 25, 2020
The main reason we switched away from New Relic is that the cost is too high. They should bring the pricing down to be more competitive. I would like to see the capacity planning improved. The security standard and compliance are areas that should be improved.
The solution needs some sort of improvement on the synthetic monitoring site. The product has good documentation for Linux, however, their documentation for Windows is lacking substantially. It's something they need to develop and mature over time. It would be great if it had out of the box integration with AWS, GCP, and other clouds. For example, if I'm using BigQuery in GCP, I want the data that I am executing in BigQuery to be reflected in New Relic.
I would like if it could have predictive analysis. Today, we only have the option to configure thresholds. In addition, it would be nice to have centralized log management, like Datadog does. As New Relic already has all of the application information and traces, it could compare them with application logs and do better analysis. Thus, it could be cheaper, have predictive analysis and log management.
They could improve the education process and how people understand that these tools are very technical. I understand everything very quickly and where it all comes in because I grew up with the product, but right now if someone was to pick it up from day one, it is a very steep learning curve. The monitoring is only as good as the alerts that it produces. By having it set up fine grain alerting, it is a bit of a pain. They already have all these other companies that use their system, so they should easily be able do alerts based on deviations that we don't need to program on a per instance or artifact basis.
Software Engineer at a computer software company with 1,001-5,000 employees
Real User
Dec 11, 2018
There is room for improvement on the alerting, insights, and infrastructure monitoring. It provides basic level alerting, and I would like more details. I would also like an infrastructure network that provides real-time views, showing the issues. Compared to their competitors, they are missing some features at the moment.
There is a picture which goes to your browser and it monitors requests from other users. However, it's impossible to use now because the price is very high. The feature is very nice, but I tried it during the trial period, and the current price makes it impossible to use.
Senior Infrastructure Architect at General Electric
Real User
Dec 11, 2018
It is complicated, especially in how you interpret the data that it provides. For someone who works in it every day, I can figure out what I want. For the general, every day developer who uses it once a month, there is large learning curve to figure out exactly the information that they want from it. If it had a bit more canned, out-of-the-box features, especially some of the reporting features, that would be more useful. Sometimes, it is difficult to work through and figure out. Some things are difficult to work through which is why I haven't done them yet, because it will take me six hours to figure out how to set them up, e.g., the dashboard. How I want it to look and how the developers might want to interpret that data, but I don't have six hours to go figure this out, and it takes a long time to do this stuff. They have this alerting capability where I can set up an alert policy, then within that alert policy, I can set up as many alerts as I want. I can set up one or I can set up a 1000. I would like a feature where I can turn off alerting at a policy level. Thus, when a policy is inactive, I can shut down all of my alerts within the policy. Right now, I have to go through them manually and deactivate each one that I don't want to use.
Senior Storage Engineer at a tech services company with 501-1,000 employees
Real User
Dec 11, 2018
I would like to have storage monitoring. E.g., being able to monitor SANS, specifically protocols, like NFS and CIFS metrics. I have not seen the ability for it to do this and having some way to do this would be awesome.
We are sharing different AWS accounts, and if a Lambda has with the same name but a different AWS account, it is a little hard to understand whether AWS or Lamda belongs to that account. Also, we have multiple accounts on the drop down to filter by Lambda, but we see two Lambda with the same name, then we don't know which one to choose. So, it needs to improve its filtering. We would like a more sophisticated filtering for the Serverless AWS pieces.
CTO and VP R&D at a tech services company with 11-50 employees
Real User
Dec 11, 2018
* I would like to be able to invest less time in IT and ad hocs. We should be concentrating on other issues. * I would like more deep dive monitoring into services and being able to install it on some apps.
VAP & IT Planning & Optimization Responsible at a comms service provider with 10,001+ employees
Real User
Oct 2, 2018
I think the APM mode can be improved. In addition, its difficult to have a predictive tool to see how the application would behave in the future when it basically only shows the historical data.
New Relic offers real-time application monitoring and insight into performance bottlenecks. Its customizable dashboards and APM integration provide efficient operational support, while server performance alerts ensure quick issue detection.New Relic provides comprehensive monitoring of application performance, tracking bottlenecks across databases and front-end components. Users employ it for server and infrastructure monitoring, as well as analyzing key metrics such as CPU and memory usage....
One thing I'd like to see in any APM, especially New Relic, is the ability to use distributed transactions. When one microservice calls another, it calls another database and microservice. The entire data visualization layer will not be able to correlate from one microservice from end to end and return on that path. Distributed transactions would be a great addition that would make life simpler. Unfortunately, no APM has that end-to-end capability. When I say "distributed transactions," I'm not only talking about the database level. It needs more and better visualization of communications across various microservices and integration with logs.
The UX/UI design of New Relic APM could be improved. The solution currently has some slow pages in terms of loading and viewing the pages, for example, the reports. The reports and other pages take a long time to load, so if that area could be improved, especially when looking for data, it would enhance New Relic APM.
Real-user monitoring would be helpful as it would help me to really understand the client-side performance of the application. Maybe for whatever reason, we have not got to explore a similar kind of feature in New Relic.
I would like the ability to set up certain dummy accounts and do the actual things that the customer is doing, without impacting the production environment. Only the read APIs are called from New Relic, not the write APIs. If we had a test account to do the write part of it, it would give us better monitoring. For example, if we are selecting the data for an existing account, we can do that part of the monitoring with New Relic. When we see failures and slowness, I would like there to be an option to do a deep dive into a collection of metrics to show the bottlenecks. It would be helpful if it didn't just state the problem, but indicate the areas to look at for a deeper resolution of the problem.
Documentation could be improved in New Relic APM, so users would have more clarity on configuring the dashboard. If New Relic gave better guidelines, users would find it easier to understand the metrics and features of New Relic APM. Another area for improvement is integration with Kubernetes. Currently, the process isn't user-friendly. It's challenging and lacks documentation for users to understand how to integrate New Relic APM with Kubernetes quickly. With multiple levels of Kubernetes dockers and other DBs on different clouds, it's tricky to gather all into New Relic APM on a single dashboard. What I'd like to see in the next version of New Relic APM is a single dashboard where you can easily view which applications fall under specific APMs. If there's a search feature where you can type in a keyword to find out if an APM is related to a particular application, that would be great.
New Relic APM could improve error debugging and the correlation with the logs. We are receiving some alerts or alarms but we need to correlate with the error log, but it is difficult if it is more than seven months retention period, it is hard to trace. We need this especially for getting historical information. Within short periods of monitoring, we do not have a problem it is only when it crosses the retention period that we are not able to correlate the errors with the log.
We would like a dashboard feature to be created for this product. This would allow us to monitor both the front and back-end of our UIs performance, and then report on it. We would also like the solution to increase the details that are provided when there are system abnormalities.
The solution is quite expensive.
I haven't come across any features that are lacking.
New Relic APM can improve the information when we dig deeper to check a problem. There should be more detailed information provided.
The solution could improve by having more network monitoring features, such as for all the infrastructure.
It offers transactions, but it does not offer an endpoint-level insight at the URL level. When we get a request, we want to know what the life cycle of that service is, and where the cycle is. This is what I am trying to locate with most of the solutions now. I am trying to research how to find a cycle per endpoint and not at the service level. It is very difficult to award the service level cycles at an endpoint level. It is important for us to get new insights to create better hygiene around the business use cases. At the endpoint level, the visibility is not that great, and metrics are not available. It gives you a full view of the entire function's execution and not from the context of the URL altogether. Also the response time, the latency contribution, and the throughput contribution are areas that need improvement. You can get the throughput contribution from New Relic, but not the latency contributions. You cannot get it at all. These are the major limitations. When working with AppDynamics, I did not find any limitations, but the same can not be said with New Relic. The way that it classifies the actual services is a bit ambiguous. It's not perfect. For example, I see there are certain solutions that are listed as extra services, as a dependency, and still I find that among load contribution, it tries to show that those services separately, which is confusing. With the transactions, when it tries to show a type of "bufferHandler" from inside, it doesn't show what the nature of the request is. Especially with Microservices, it doesn't show what kind of method is present, which makes finding data very difficult. Instead, you need to go to the raw data. I think that defeats the purpose of using this tool. The transactions do not show the time consumed by the request, from the metrics execution perspective. It was suggested that I did not know how to read it but I have done all that I could. It is very difficult to relate to and requires a lot of experience and time to read through, which it should not. It should not be difficult to find the latency and throughput for the entire system when requested. It should not be difficult to develop the data that relates to the various types of execution. It should have complete exposure around the endpoints. The services-to-service dependency is fine but most of the startups have only one or two services that are all cycled. It does not provide you with a lot of help when you are showing that the two services are dependent. What all of the dependent endpoints are and how are the cycles being formed is information that should be available in most tools, but not with New Relic and some other tools.
I haven't yet had too much of a chance to read up on the product or to compare it to other solutions to see if there are items that are standard elsewhere but lacking in this product. I need more time with the solution to really talk about any shortcomings if they exist at all. It would be nice if there were pre-made dashboards. There should be out-of-the-box options that you can have. It would be nice to have something already done, so that you don't need to reinvent the wheel again, to create your own dashboard. The pricing might be high if you are a company that goes through a lot of data.
There has been some problem with the agent, and it is just not working well. It is not able to record information with the application server. They have been able to fix the issue, but it took quite a long time. This is the main issue in the APM products and also in New Relic. The mobile application monitoring has been pretty difficult to set up and also quite expensive. It should be a little bit easier and cheaper. Because it is pretty difficult and expensive, many customers don't take it.
There really is nothing that stands out with New Relic. With the insight, I think it will be found lacking for its report aggregation capabilities. How granular I could go down at looking at certain data, especially related to the operations, is limited. The API integrations that they have for us to automate our configuration was fine, but I think for some of these tools, it was over-engineering for us to try and automate any of that. So, we just use the user interfaces.
The older view is much better than the new view that they have. We'd like to go back to that previous version. The user interface just isn't as nice as it used to be. The solution does not seem to be lacking any features. I can't say that I recommend any specific features for the next release.
The main reason we switched away from New Relic is that the cost is too high. They should bring the pricing down to be more competitive. I would like to see the capacity planning improved. The security standard and compliance are areas that should be improved.
The solution needs some sort of improvement on the synthetic monitoring site. The product has good documentation for Linux, however, their documentation for Windows is lacking substantially. It's something they need to develop and mature over time. It would be great if it had out of the box integration with AWS, GCP, and other clouds. For example, if I'm using BigQuery in GCP, I want the data that I am executing in BigQuery to be reflected in New Relic.
Data retention should be increased, advanced RCA and all possible and best recommendations should be provided. Pricing should be revised
I would like to see the company implement the AI auto-baseline feature which Dynatrace has.
I would like if it could have predictive analysis. Today, we only have the option to configure thresholds. In addition, it would be nice to have centralized log management, like Datadog does. As New Relic already has all of the application information and traces, it could compare them with application logs and do better analysis. Thus, it could be cheaper, have predictive analysis and log management.
They could improve the education process and how people understand that these tools are very technical. I understand everything very quickly and where it all comes in because I grew up with the product, but right now if someone was to pick it up from day one, it is a very steep learning curve. The monitoring is only as good as the alerts that it produces. By having it set up fine grain alerting, it is a bit of a pain. They already have all these other companies that use their system, so they should easily be able do alerts based on deviations that we don't need to program on a per instance or artifact basis.
There is room for improvement on the alerting, insights, and infrastructure monitoring. It provides basic level alerting, and I would like more details. I would also like an infrastructure network that provides real-time views, showing the issues. Compared to their competitors, they are missing some features at the moment.
The APIs could be better. I would also like more APIs and features to integrate with streaming solutions, like Kinesis or Kafka.
There is a picture which goes to your browser and it monitors requests from other users. However, it's impossible to use now because the price is very high. The feature is very nice, but I tried it during the trial period, and the current price makes it impossible to use.
It is complicated, especially in how you interpret the data that it provides. For someone who works in it every day, I can figure out what I want. For the general, every day developer who uses it once a month, there is large learning curve to figure out exactly the information that they want from it. If it had a bit more canned, out-of-the-box features, especially some of the reporting features, that would be more useful. Sometimes, it is difficult to work through and figure out. Some things are difficult to work through which is why I haven't done them yet, because it will take me six hours to figure out how to set them up, e.g., the dashboard. How I want it to look and how the developers might want to interpret that data, but I don't have six hours to go figure this out, and it takes a long time to do this stuff. They have this alerting capability where I can set up an alert policy, then within that alert policy, I can set up as many alerts as I want. I can set up one or I can set up a 1000. I would like a feature where I can turn off alerting at a policy level. Thus, when a policy is inactive, I can shut down all of my alerts within the policy. Right now, I have to go through them manually and deactivate each one that I don't want to use.
I would like to have storage monitoring. E.g., being able to monitor SANS, specifically protocols, like NFS and CIFS metrics. I have not seen the ability for it to do this and having some way to do this would be awesome.
We are sharing different AWS accounts, and if a Lambda has with the same name but a different AWS account, it is a little hard to understand whether AWS or Lamda belongs to that account. Also, we have multiple accounts on the drop down to filter by Lambda, but we see two Lambda with the same name, then we don't know which one to choose. So, it needs to improve its filtering. We would like a more sophisticated filtering for the Serverless AWS pieces.
We have had issues with our agents going offline.
The price needs improvement.
* I would like to be able to invest less time in IT and ad hocs. We should be concentrating on other issues. * I would like more deep dive monitoring into services and being able to install it on some apps.
We would like to receive more AWS-specific details from the New Relic Dashboard, like EC2 health.
I think the APM mode can be improved. In addition, its difficult to have a predictive tool to see how the application would behave in the future when it basically only shows the historical data.