What needs improvement with Sonatype Nexus Lifecycle?

While Sonatype Lifecycle effectively manages artifacts in Nexus Repository and performs code firewall checks based on rules, it has the potential to expand further. I am looking forward to additional features similar to SonarQube, especially since licenses are often split per component. SonarType could integrate cloud-based capabilities, addressing the increasing shift towards cloud workloads. While there have been demos and discussions around this, significant progress on scanning and analyzing cloud images remains to be seen. I am looking forward to Sonatype incorporating these enhancements, particularly in regard to cloud-based features. On-prem workloads are getting to the cloud workloads. * I would like to see more cloud-related insights, such as logging capabilities for the images we use and image scanning information. * Additionally, it would be beneficial to have insights into the stages of dependencies and ensure they comply with standards. If there are any violations in respect to CVSS reports, * Integrating CVSS (Common Vulnerability Scoring System) report rules into the Lifecycle module to detect and report violations would be valuable. I am hoping to see these enhancements from Sonatype in the future. On the security side, I think there's a lot of development needed. There are many security tools on the market, like open-source ones, that Sonatype doesn't integrate with.

Improvements are needed as per customer requirements.

There is room for improvement in the code analysis aspect of Sonatype Lifecycle, specifically in the area of deployment security. While the product effectively scans components and provides threat intelligence, it requires additional manual effort to ensure that the configuration of the product during deployment is done securely. When it comes to new features, I would find it incredibly beneficial if Sonatype Lifecycle could integrate with deployment tools, enabling real-time identification of any vulnerabilities as developers push code to production.

One downside to it is that it is costly. I can see it only for enterprises. I cannot see it for small businesses or for individual use. The configuration part is a little bit tricky. There is a learning curve there because it has multiple components. If someone has used another type of scanner, they would not think of the configuration intuitively. The configuration part can be better. Installation is straightforward, but the configuration can be better. It can be improved. There is a learning curve. Before we started using this tool, I did a lot of sessions with the vendors themselves to give an overview to the people. I also did a small documentation on how to install it because there are many components here and there. You need to understand how everything is put together. They can integrate it or make it a simpler process. During the short experience that we have had with it, we have noticed that some of the languages such as JavaScript and TypeScript consume high resources. They take a longer time to scan. Memory consumption is also very high for those languages. We are working with Fortify to find ways to optimize the scan. I noticed this with these types of languages. By nature, they take time. It can be tricky if you want to exclude some files from scanning. For instance, if you do not want to scan and push testing files to Fortify Software Security Center, that is tricky with some IDEs, such as IntelliJ. We found that there is an Exclude feature that is not working. We reported that to them for future fixing. It needs some work on the plugins to make them consistent across IDEs and make them easier. For integration with IDEs, they have so many plugins. For example, they have something called security analysis, and they have something called remediation. As a user, I would love to have them as one. Why should we have two plugins in the same IDE? Just give me one plugin that I can hook to the tool and use it. This is one thing. Some of the features in these plugins also need more testing. They are not consistent across all the IDEs. From what I saw, there are different options in these tools. For example, if you install it with IntelliJ, it will be different from VS Code. Some options are different, or one tool has more options than others. They can invest more in making them consistent.

It would be nice if they had a version suitable for single developers that could be more cost-effective and maybe faster to learn.

Fortify's software security center needs a design refresh. It has maintained the same design for the entirety of our five years of use, making it feel outdated compared to its FOD portal, which receives regular bi-monthly updates. This area is a prime candidate for improvement in the future. Fortify needs to move to a more frequent release cycle Currently, they only release two updates per year, which is considerably slower than their peers, so I would very much like to see that improve.

The price can be improved.

Not all languages are supported in Fortify. They should expand their language offering.

Fortify Static Code Analyzer has a bit of a learning curve, and I don't find it particularly helpful in narrowing down the vulnerabilities we should prioritize. It throws everything at us at once, which can be overwhelming. While it's not a major issue, I'd like to see it focus on critical vulnerabilities and highlight them upfront. Furthermore, categorizing critical vulnerabilities by platform-specific vulnerabilities and relevance to supported features would be incredibly beneficial. While Fortify Static Code Analyzer has some merit, I believe it still has significant room for improvement. We have encountered a high number of false positives, which has been a major obstacle and resource drain.

One downside to it is that it is costly. I can see it only for enterprises. I cannot see it for small businesses or for individual use. The configuration part is a little bit tricky. There is a learning curve there because it has multiple components. If someone has used another type of scanner, they would not think of the configuration intuitively. The configuration part can be better. Installation is straightforward, but the configuration can be better. It can be improved. There is a learning curve. Before we started using this tool, I did a lot of sessions with the vendors themselves to give an overview to the people. I also did a small documentation on how to install it because there are many components here and there. You need to understand how everything is put together. They can integrate it or make it a simpler process. During the short experience that we have had with it, we have noticed that some of the languages such as JavaScript and TypeScript consume high resources. They take a longer time to scan. Memory consumption is also very high for those languages. We are working with Fortify to find ways to optimize the scan. I noticed this with these types of languages. By nature, they take time. It can be tricky if you want to exclude some files from scanning. For instance, if you do not want to scan and push testing files to Fortify Software Security Center, that is tricky with some IDEs, such as IntelliJ. We found that there is an Exclude feature that is not working. We reported that to them for future fixing. It needs some work on the plugins to make them consistent across IDEs and make them easier. For integration with IDEs, they have so many plugins. For example, they have something called security analysis, and they have something called remediation. As a user, I would love to have them as one. Why should we have two plugins in the same IDE? Just give me one plugin that I can hook to the tool and use it. This is one thing. Some of the features in these plugins also need more testing. They are not consistent across all the IDEs. From what I saw, there are different options in these tools. For example, if you install it with IntelliJ, it will be different from VS Code. Some options are different, or one tool has more options than others. They can invest more in making them consistent.

The reporting could be better.

It could be because I need to learn more about Sonatype Nexus Lifecycle, but as a leader, if I want to analyze the vulnerability situation and how it is and the forecast, I'd like to look at the reports and understand what the results mean. It's been challenging for me to understand the reports and dashboards on Sonatype Nexus Lifecycle, so I'll need to take a course or watch some YouTube tutorials about the product. If Sonatype Nexus Lifecycle has documentation that could help me properly analyze the vulnerability situation and what the graphs mean, then that would be helpful. I need help understanding what each graph is showing, and it seems my company is the worst, based on the chart. Still, I need clarification, so if there were some documentation, a more extensive knowledge base, or a question mark icon you could hover over that would explain what each data on the graph means, that would make Sonatype Nexus Lifecycle better.

Sonatype Nexus Lifecycle can improve by having a feature to automatically detect vulnerabilities. Additionally, if it could automatically push the dependencies or create notifications it would be beneficial.

The main drawback of this product is that it's not an SaaS solution and they really need to build a complete SaaS product. Although the vulnerability detection accuracy is good, the solution is quite weak when it comes to remediation accuracy which is not good. They are currently sorting by component versions and the sorting algorithm is not correct, it requires a proper tool.

I'm not using the technology directly, and I haven't heard anything from our customer base. As far as I know, Sonatype has a unique customer engagement framework with a regular customer meet-up to go through deployment issues. They take feedback directly from the customer.

Sonatype Nexus Lifecycle can improve the functionality. Some functionalities are missing from the UI that could be accessed using the API but they are not available. For example, seeing more than the 100 first reports or, seeing your comments when you process a waiver for a vulnerability or a violation. When you submit a waiver, you enter a comment, and when you need to access this comment, in the reports, you don't see it. This is a drawback.

We got a lot of annotations for certain libraries when it comes to Java, but my feeling, and the feeling of a colleague as well, is that we don't get as many for critical libraries when it comes to .NET, as if most of them are really fine. It's true that we have more Java applications than .NET, but the number of our applications in the .NET area will increase. Again, it's just an impression, but it seems that the annotations for .NET are not the same as for Java. It would be good if Sonatype would check the status of annotations for .NET packages. Again, I note that we are just starting to use an open source library from NuGet for the .NET area, while we have been using it for Java for several years and we are using more packages. For .NET, it's evolving. But my impression is that annotations are more focused on Java, and that in .NET we just do not see as many security issues as in Java. It could be fine, but maybe Sopatype started with Java and then expanded the portfolio to .NET and to other languages. This is something which could be further checked. It could also be the fact that we have had Java applications for around 20 years, using open source libraries. When you go to the newer versions, you need to check and test. Whereas the .NET applications are evolving and are using open source libraries, and the .NET side is really new for our organization.

The user interface needs to be improved. It is slow for us. We use Nexus IQ mostly via APIs. We don't use the interface that much, but when we use it, certain areas are just unresponsive or very slow to load. So, performance-wise, the UI is not fast enough for us, but we don't use it that much anyway.

One area of improvement, about which I have spoken to the Sonatype architect a while ago, is related to the installation. We still have an installation on Linux machines. The installation should move to EKS or Kubernetes so that we can do rollover updates, and we don't have to take the service down. My primary focus is to have at least triple line availability of my tools, which gives me a very small window to update my tools, including IQ. Not having them on Kubernetes means that every time we are performing an upgrade, there is downtime. It impacts the 0.1% allocated downtime that we are allowed to have, which becomes a challenge. So, if there is Kubernetes installation, it would be much easier. That's one thing that definitely needs to be improved. Some of our engineers came from outside of BT, and there are some features that they are used to from rival products, but they are currently not there in Sonatype IQ. For example, Snyk has a feature to stop a particular check-in from happening at the merge stage in case something is different or wrong. This feature is still in the development phase in IQ. Such a feature would be handy in IQ. Another area where Nexus can severely improve is the licensing model. I am not worried about the licensing cost, but the way they calculate the number of licenses being used needs to be improved. They have been quite ambiguous in terms of how they calculate who is using Nexus or IQ, and this ambiguity has not been good. At times, we think we have a certain number of customers, but Sonatype says that it is not true, and we have some other number. They haven't been able to explain very well how they calculate that number, which has been a challenge for us.

The integration is one sore spot, because when we first bought the tool they said JavaScript wasn't really part of the IDE integration, but it was on the roadmap. I followed up on that, and they said, "Oh, you can submit an idea on our idea site to have that added." The sales team said it was already in the pipeline, but it was actually not in the pipeline. Overall it's good, but it would be good for our JavaScript front-end developers to have that IDE integration for their libraries. Right now, they don't, and I'm told by my Sonatype support rep that I need to submit an idea, from which they will submit a feature request. I was told it was already in the pipeline, so that was one strike against sales. Everything else has been pretty good. Also, when Nexus Firewall blocks a component, it doesn't really give us a message that tells us where to go; at least it doesn't in our setup. I have to tell all the users, "Here's the URL where you can go to look up why Firewall is blocking your stuff. And that is odd because when it finishes a scan, the scan results give you the URL. But when you get blocked by Firewall, it doesn't give you the URL where you can go look that up. You can definitely work around that, but it's a bit strange. It's almost like something they forgot to include.

Nexus Lifecycle is multiple products. One drawback I've noticed is that there are some differences in the features between the products within Lifecycle. They need to maintain the same structure, but there are some slight differences. Other than that, the tool is very user-friendly and gives the right reports to the right teams.

One thing that it is lacking, one thing I don't like, is that when you label something or add a status to it, you do it as an overall function, but you can't go back and isolate a library that you want to call out individually and remove a status from it. It's still lacking some functionality-type things for controlling labels and statuses. I'd like to be able to apply it across all of my apps, but then turn it off for one, and I can't do that. I have to go to all 100 apps and do it individually in order to get something on each one, and I don't like that. I should be able to add it as a group and remove it as a single. Everything else has been really good.

The thing that they're already addressing is high-quality data for the Conan dependencies. They're very responsive to user needs. We're one of the first organizations to use Conan, so I identified a discrepancy in how they were scanning the dependencies, and they added the functionality within four weeks or so. The team is incredible. I can't think of any other ways that it could improve it. When Conan support was first added to Nexus IQ, it would only scan one file type for dependencies. We don't use that specific method in Conan, but rather, another acceptable method for declaring dependencies that IQ wasn't scanning. I think the Sonatype developers didn't even know about it because they learning Conan as much as we were. I informed them of the other file type for declaring dependencies and they quickly added the functionality.

The reporting capability is good but I wish it was better. I sent the request to support and they raised it as an enhancement within the system. An example is filtering by version. If I have a framework that is used in all applications, but version 1 is used in 50 percent of them and version 2 in 25 percent, they will show as different libraries with different usage. But in reality, they're all using one framework.

One thing that I would like to give feedback on is to scan the binary code. It's very difficult to find. It's under organization and policies where there are action buttons that are not very obvious. I think for people who are using it and are not integrated into it, it is not easy to find the button to load the binary and do the scan. This is if there is no existing, continuous integration process, which I believe most people have, but some users don't have this at the moment. This is the most important function of the Nexus IQ, so I expect it should be right on the dashboard where you can apply your binary and do a quick scan. Right now, it's hidden inside organization and policies. If you select the organization, then you can see in the top corner that there is a manual action which you can approve. There are multiple steps to reach that important function that we need. When we were initially looking at the dashboard, we looked for it and couldn't find it. So, we called our coworker who set up the server and they told us it's not on the dashboard. This comes down to usability. There is another usability thing in the reports section. When the PDF gets generated, it is different from the web version. There are some components from some areas which only reside inside the PDF version. When I generate the PDF for my boss to review, she comes back with a question that I didn't even see. I see on the reporting page whatever the PDF will be generating. The PDF is actually generating more information than the web version. That caught me off guard because she forwarded this to the security officer, who is asking, "Why is this? Or, why is that?" But, she has no idea. I didn't have anything handy because I saw the PDF version, which should be same as what I see on the web. This is a bit misrepresented. I would like these versions to speak together and be consistent. Printing a PDF report should generally reflect whatever you have on the page.

The biggest thing that I have run into, which there are ways around, is being able to easily access the auditing data from a third-party tool; being able to pull all of that into one place in a cohesive manner where you can report off of that. We've had a little bit of a challenge with that. There are a number of things available to work with, to help with that in the tool, but we just haven't explored them yet.

Some of the APIs are just REST APIs and I would like to see more of the functionality in the plugin side of the world. For example, with the RESTful API I can actually delete or move an artifact from one Nexus repository to another. I can't do that with the pipeline API, as of yet. I'd like to see a bit more functionality on that side.

We do not use it for more because it is still too immature, not quite "finished." It is missing important features for making it a daily tool. It's not complete, from my point of view. All the Lifecycle tools are not yet finished and usable in production. We are using it in production, but it's not fulfilling all our needs. It's not yet finalized. It's the right kind of tool and going in the right direction, but it really needs to be more code-driven and oriented to be scaled at the developer level. Nowadays, developers need to be autonomous, so we need to be able to supply them tooling and then everything should be orchestrated around the principle of GitHubs. That is really important in IQ because developers will want to make changes and they need things very quickly. Everything should be driven by that but that is not yet the case. The features are interesting but the way those features are configured and tuned is not quite there yet. There is room for improvement in the way it is managed, having code-driven configuration, and automation. It needs not to be an old-style tool. Today, a computer tool must be usable in many ways: as a client for developers, as a webpage and reporting tool for managers, and as an automated blocker for continuous integration. It must have a REST API and it must have many features that make it usable in many ways. Currently, that's not the case. One thing I can say that is very positive is that it is much better than the other tools. But regarding usage, it's not perfect. It's missing everything around the tuning and usage.

One of the things that we specifically did ask for is support for transitive dependencies. Sometimes a dependency that we define in our POM file for a certain library will be dependent on other stuff and we will pull that stuff in, then you get a cascade of libraries that are pulled in. This caused confusing to us at first, because we would see a component that would have security ticket or security notification on it and wonder "Where is this coming in from?" Because when we checked what we defined as our dependencies it's not there. It didn't take us too long effort to realize that it was a transitive dependency pulled in by something else, but the question then remains "Which dependency is doing that?" This is the biggest thing that we have talked to Sonatype about. Even though we have found an way to see where transitive dependencies are coming from, it would be nice if this was visible through IQ Server as well. Another issue is that, although Sonatype categorizes and indexes a lot of different repositories, it doesn't index every single repo in existence. One of the components we used switched where it came from, so a later version was actually coming from a different repository that Sonatype didn't index, as it was relatively smaller. They cover a large amount of available libraries, but they don't cover 100 percent of them. In this case, that component that was marked as an unknown component. When we get this kind of notification, we have to double check it. That is how we found out that these are components aren't covered by Sonatype yet. We have put in requests to have this particular repository added to the sources Sonatype indexes. It's something to be aware of if you use obscure repositories.

It would be helpful if it had a more detailed view of what has been quarantined, for people who don't have Lifecycle licenses. Other than that, it's pretty good.

We use Azure DevOps as our application lifecycle management tool. It doesn't integrate with that as well as it does with other tools at the moment, but I think there's work being done to address that. In terms of IDEs, it integrates well. We would like to integrate it into our Azure cloud deployment but the integration with Azure Active Directory isn't quite as slick as we would like it to be. We have to do some workarounds for that at the moment. Also, the ability of the solution to recognize more of the .NET components would be helpful for us.

The GUI is simple, so it's easy to use. It started as great to use, but for larger scale companies, it also comes with some limitations. This is why we tried to move to more of an API approach. So, the GUI could use some improvements potentially. Something else that's a bit lacking is most of our components are not explicitly included but are transitive dependencies. We have 50 applications that all report security issues, but they all come from one central library that we built ourselves, which is also scanned by Lifecycle. So, we have 51 components, and we are not seeing that only one of them is really the one we should be targeting. What would be really great in the solution would be some dependency graphing, or at least collecting the transitive dependencies. That would help for larger scale implementations. The Success Metrics report is really focused on very specific numbers that are not interesting to us. They are for when you are much further along in the onboarding process. There is an API which allows you to retrieve the data on which the Success Metrics are based. We use this API to create our own charts, reflecting what we're looking for.

We use Griddle a lot for integrating into our local builds with the IDE, which is another built system. There is not a lot of support for it nor published modules that can be readily used. So, we had to create our own. No Griddle plugins have been released. One of the challenges is getting the policy correct. You need to understand when to grandfather components, then come back and do it. Currently, there's no feature in Nexus IQ which says when you grandfather a component, or behave a component. There's no feature to remind me again in two months' time, for example. I had to access a grandfather competent today because I couldn't afford to fix it because of different constraints. I might grandfather it for now, or I might leave it for now, but if there was an option to remind me in two months, or unwaive it in two months' time, that would make it seamless. That way I wouldn't have to remember that there's something to be done. It would automatically start breaking bills and automatically someone will look at it.

They have recently released some online training documentation, because we had to a lot of our own learning. If they had a more comprehensive online tutorial base, both for admin and developers, that would help. It would be good if they actually ran through some scenarios, regarding what happens if I do pick up a vulnerability. How do I fork out into the various decisions? If the vulnerability is not of a severe nature, can I just go ahead with it until it becomes severe? This is important because, obviously, business demands certain deliverables to be ready at a certain time. So, some online tutorials around the administration, around developers, on how to use the tool effectively, would be a good idea. Where we have extended is that we're using third-party Docker containers. So, we're not just using third-party libraries, but we also have third-party-type Docker containers, or containers from Docker Hub, for example. Somebody else has built the Docker image and we're using the Docker image. Scanning of security and vulnerabilities on that image, specifically, would be useful. It would be good, as we're building an image, or as we're running an image, if we could decompile that image and scan it, to look for any vulnerabilities or any areas where there's been a violation of licensing. For example, we could download a WebLogic container, which could be an infringement of licensing. It's things like that which our developers need to be mindful of. They could simply download any container from Docker Hub, without being aware of the licensing violations or the security vulnerabilities.

Overall, it's pretty good. The drill-through and search capabilities are pretty good, they're not horrible. As far as the relationship of, and ease of finding the relationships between, libraries and applications across the whole enterprise goes, it still does that. They could make that a little smoother, although right now it's still pretty good. It's taking an eight out of ten and asking it to be a ten.

It doesn't provide real-time notifications from the scans. We have to re-scan every time, whenever a build happens. Also, since Nexus Repository just keeps on adding the .jar artifacts whenever there is a build, whenever an application is going up, there is always a space issue on the server. That is one of the things that we are looking for Nexus to notify us about: if it is running out of space.

Getting it integrated depends on your structure and how your DevOps teams are structured. The biggest thing is getting it used uniformly across all the different teams. It's more of a process issue. The process needs to be thought out about how it's going to be used, what kind of training there will be, and how it's going to be socialized, how it's going to be rolled out and controlled, enterprise-wide. That's probably more of a challenge than the technology itself. It's pretty simple to get up and running. It's not really an enterprise solution, like Active Directory, which you can enforce on everyone. It's something that's done through each little vertical.

We created a Wiki page for each team showing an overview of their outstanding security issues because the Lifecycle reporting interface isn't as intuitive. It is good for people on my team who use it quite often. But for a tech engineer who doesn't interact with it regularly, it's quite confusing. We did that because we got so many questions about it all the time. There are other areas for improvement. The most recent one - something I haven't shared with Sonatype yet but I intend to - is with the creating of defect tickets. The solution has something that is really useful, its integration with JIRA, and it creates tickets if there's an issue. What I thought would be really good was, from the moment we break builds, there is no way to track, from a management perspective, how we are doing. We are looking at creating tickets. The problem with the tickets, which is the where there is room for Sonatype to grow, is that there is no flexibility in terms of customizing the entries in the tickets. There are certain things they put in for you, they tell you what application it is, but what I'd really like to be able to do is say, "Fill in this field with the name of the application. Fill in this field with the name of the owner. Or set a due date to be X days from when it was raised. They don't allow that. They allow hard-coded values across everything in Nexus IQ. It doesn't work well because the tickets created depend on the use case. We would like to create these tickets and give them directly to the teams that have to look after them. We want to be able to assign them to the right person, based on the application that is used. " We are looking at finding ways to integrate with it because they don't have that. Another feature they could use is more languages. Sonatype has been mainly a Java shop because they look after Maven Central. And we have been mainly a Java shop in development. But we've slowly been branching out to different languages. They don't cover all of them, and those that they do cover are not as in-depth as we would like them to be. They don't have the same level of coverage as the main language, which is Java.

If there is something which is not in Maven Central, sometimes it is difficult to get the right information because it's not found. And if you look at NPM-based applications, JavaScript, for example, these are only checkable via the build pipeline. You cannot upload the application itself and scan it, as is possible with Java, because a file could change significantly, so the applications are not found anymore. This is something that could be improved in future. Also, I have seen in Black Duck, for example, that there is also information about exploits there are known for a given vulnerability. This is something I haven't seen or haven't found yet in Nexus Lifecycle. If there is a known exploit to a vulnerability, this could be something that is useful to know as well.

The only thing I can say is that sometimes we face difficulties with Maven Central. We are integrating everything with that, as a repository. If Maven Central changes something in its versions... For example, if I'm using the 1.0.0 version, after one or two years, the 1.0.0 version will be gone from Maven Central but our team will still be using that 1.0.0 version to build. When they do builds, it won't build completely because that version is gone from Maven Central. There is a difference in our Sonatype Maven Central. That is the only issue I have seen so far. If an old version is gone, it's not able to use it anymore. Is there any way we can keep the old versions in our local repository instead of in Maven Central?

They could do with making more plugins for the more common integration engines out there. Right now, it supports automation engine by Jenkins but it doesn't fully support something like TeamCity. That's where they could make the most improvements. In terms of features, the reports natively come in as PDF or JSON. They should start thinking of another way to filter their reports. The reporting tool used by most enterprises, like Splunk and Elasticsearch, do not work as well with JSON. They should improve the reporting so that the format can be expanded.