Aternity AppInternals [EOL] Reviews

The application side of the solution is its most valuable aspect. We are using it on the logs that were captured on the application side. 

As an Administrator, before we bought this AppInternals, I didn't have visibility on why items were slow or why an application was not running. This gives us the ability to see what's going on. The application is load balancing. We can now see if its own server has issues or just one specific server has issues.

Right now, our challenge is that we don't have the support we need, and based on the requirements we have, there are some audits and it turns out that the AppInternals agent is causing false trafficking issues.

We could not update the server requirements on the version we have right now, due to the support. 

It would be great if the solution could offer fixed bundles and more features.

We have been working with the solution for three or four years now. 

It's not updated to the latest version, due to the fact that if we do that it will crash or cause some issues. Because we don't have a new AP version, we have just direct production. If our management accepts the risk, then we might do the upgrade of the server as well as its application version. Currently, we are using a two-year-old version of the solution. It is not up to date.

The server-based scalability is a challenge, however, we cannot change anything, including the scale, unless the management decides to take the risk and update the solution.

We're running it in a Linux bot. I'm not sure if I would describe the process as complex. On our side is basically in-house, so it's a pretty standard installation process. Whatever servers that we have provide talks to the app servers which talk to the database servers. The setup is intranet-based and not internet-based. That makes our setups rather simple and straightforward.

We handle everything in house using our own team.

I don't know the exact pricing, however, my understanding is that it's quite expensive. Dynatrace has competitive pricing in comparison. I believe Cisco pricing is also okay.

We recently increased to another license. That said, for other teams to be involved in what we do, including network and infrastructure, we've decided to look into Dynatrace, which is a more complete system.

We are looking at Dynatrace based on some information that we've gotten from websites as well as some other experiences from our colleagues. Our manager also has previous experience on the solution. We're now in the POC stage and are awaiting the arrival of the hardware to get started.

Before this product, I recommended the Sophos Manage Engine. Management did not consider it because it's from India.

We're customers. We don't have a special relationship with the company. I am the operation's application administrator.

At the moment, the solution is monitoring the applications that we want it to monitor. We're not looking at it on a daily basis. We just use it here and there. The challenge we have is that we believe that Dynatrace can address some tasks we'd like done that are not possible to execute on this solution. Dynatrace includes monitoring of not just the logs, but the applications as well.

Dynatrace also makes us include firewalls, etc. We could have it so there are end to end solutions or identifications within the process. What we have right now is only the app server as well as the database server and its connectivity.

The solution does save time that's caused when creating SQL coding. However, there are some issues around AppInternals changing the code and giving us issues.

From the point of view of the top management, we want a more permanent solution. We want to identify what causes issues and we are looking for a more end-to-end solution. That is why we are looking at Dynatrace. 

I don't know if I have any advice for other organizations. The experience I have is only based on my own limited experience. However, I believe that the application has the potential to be really good if it identifies what is really important to an organization. It's still sort of half-baked without AppResponse. Without the other key features in infrastructure, like the network, firewalls, and proxies, it's lacking. It's not fully complete.

I'd rate the solution six out of ten. It addresses some of our needs, but there's a lot lacking that could be added to make it a fuller, more interesting product.

The most valuable aspect of this solution is the integration with their other systems. It's easy to understand and it points out the relevant problems in the enterprise.

I would like for it to have automated updates, the way the product updates itself should be all automated, as opposed to what it is now. 

I have been using Riverbed for several years now. 

The stability is good.

Scalability could be improved in terms of the number of technologies that you can use it with.

The people using is a small amount, just our enterprise. The users are mostly admin.

Their technical support is good.

We used Foglight before this solution. The reason we switched is because Quest Foglight was a bit of a complex platform and not really delivering the visibility that was required.

The initial setup is not complex, but it's not as smooth as it could be. 

The deployment takes a couple of days, but it's the ongoing deployment of the agents that's the issue.

I would rate it an eight out of ten. 

My advice to someone considering this solution is to do a proof of concept. Clearly identify what it is you actually want a product of this nature to do. How grandly do you want to get into the data sets?

We primarily use this solution for troubleshooting. Whenever a customer has problems in the performance with the servers, we use this application to manage this do some troubleshooting to fix the problem.

The most valuable feature is the connections for marketing. I like it that one can match IPs with the application name.

I would like to see an improvement in the investigation. Sometimes one has some issues or something like an error that needs to be investigated to take appropriate action. But I have found a server error that I can't seem to move. 

I also think the support could be better. They weren't very helpful when I had a problem. 

The technical support is not very good and should be improved.

My advice to others would be to use this solution for monitoring. For instance, if I need to monitor something or if I have an error, I can take access from Riverbed, which will then provide me with a reports of what I should do. That is very handy. One should really understand how the solution works in order to get the best out of it. You can use it in many ways - you can do troubleshooting, make reports, you can read, and it will really help you, but you need to deliver it in the right way.

I will rate it a seven out of ten. 

Cuts down on the time to resolve complicated application issues. We have been able to trace individual transactions to identify where the issue is whether in the application code, server or network.

Transaction Tracing is the most useful. Being able to have the transaction stitched together so we can see where the problem is has proven invaluable.

Support for PHP, DB and other applications need to be supported. We have a broad range of applications that we would like visibility into but the list of currently supported applications is very small.

We had no issues with stability of either the agent or console.

We don’t have any scalability issues but we also do not have a large install base, yet.

9/10 - we had a couple of issues that required opening up tickets. The Riverbed support engineers usually resolve issues in a timely manner.

I didn't use anything similar to this in the past.

The initial setup was pretty straightforword. We watched the installation videos and read the documentation to prepare. The two issues we ran into were permissions of the installation account to the folder structure and the agent not being supported on particular OS versions.

It’s pricey. The licensing changed from v9 to v10. The licensing model for v9 is better where it is an individual license per server, while v10 licenses are per JVM/ .NET and server instance. The latter model appears to be the model that other APM vendors are using.

This solution was chosen prior to my joining the company.

Read the documentation and watch as many videos as you can. Also verify that your application servers OS support the agent.

I was part of a Test Organization (focused on Performance Engineering) and TTW allows our engineers to provide many more improvements opportunities to the developers while improving software performance. It was not just saying where the problem was, but the tool was an enabler for more technical people to nail down the root cause of problems and propose better solutions. After the tool was installed in applications, the confidence of the team improved towards raising more specific performance defects.

This product has really wonderful features and one of the most valuable ones is the transaction searches into TTW. This will show you the operations that occurred under conditions that the user specifies, giving you the opportunity to drill down on the things that are more important and affecting system performance.

Besides that, the capability of analysing each individual transaction captured to a very low level detail (method call/line of code). This ability empowers the testing teams to provide more visibility to developers on how they can improve.

There are some functional bugs when using TTW and making report queries (to generate graphs). I remember them failing once in a while and then the time taken to complete a task was higher since long running reports needed to be played again.

Also, they should find a way for report generation from TTW to run quicker. Maybe it is the case of telling the user what kind of indexes/options could be used in each type of report to make it faster (in a help or “hint” approach).

The implementation we had in our organization was very stable. I

I would say there were opportunities for scalability, since many engineers were triggering heavy reports it took time to complete. However, as I was not managing infrastructure, I am not sure if we have used all resources.

I have not used technical support.

I have worked, in the past, with Mercury Diagnostics for performance troubleshooting and later (currently) with AppDynamics. From the experience I had with these three tools, I realized that Riverbed was superior in all aspects (performance, user experience, and features).

Since it was not me doing this activity, I cannot comment on it.

When I worked with Riverbed, there was an internal team responsible for implementation and maintenance. I took the user role to provide input to development teams.

I have not participated in negotiations. Cannot comment.

We really like the dashboards and their latest release. By going to v10 it made our lives a whole lot easier because prior to v10, we had to put all of these assets, these Riverbed arguments, basically JVM start up itself. Then when they send out a new release or a new patch on it, the library changed to 64-bit. If that library changed, we have to go to all the different JVMs we have implemented and we would have to go in there and change all of the start-up arguments. When we go with system wide enabling, we don't have to mess around with the JVM arguments anymore. We just control on the backend of AppInternals what we want to instrument and what we don't want to instrument. That makes it a lot easier.

One thing that I can't do with the AppInternals - we have put in a request as an enhancement that we would like to search for items. If we're searching for a URL and we want to know was it a get or was it a post.

Those are attributes that are of interest to us because we can have a static URL but rather than just seeing the performance time, the metrics and everything around that, we needed a way to go in and carve those out. To find out which one was a get and which one was a post. We would like to see those columns. We would like to have those columns available that the customized columns had.

With v10, Riverbed has done a very good job and we've been using it a lot. There's actually more features in v10 that we really haven't gotten into yet. I'm planning on making some more time to dig into those.

In terms of stability we haven't had any issues there. There was one time when we were trying out different versions and when we tried v9, for some reason it didn't working well. We then actually would go ahead and hang up our transactions, where they wouldn't be able to go through the application.

That was something that they corrected. With the later part of version nine, we went ahead and rolled it out and the problem was corrected. We did see that little bump in the road where we caused a little bit of aggravation. We always try things out in our non-prod environment first to see how they are going to work. The good thing is we didn't have it in production yet. We were able to catch this anomaly, happening in the non-product. Riverbed engaged, they were are extremely responsive and they got that corrected for us. Apart from that one little bump in the road there, you know that we've been happy with it. We have been happy with it and it's been pretty stable. It has not caused any issues.

This team here is definitely scalable. Right now we have about 100 licenses in prod and 100 licenses in non-prod and we are not using all of them yet, but we are getting pretty close. In terms of scalable within the organization, we do have other groups that can start using them more. Right now, I think we are mature enough with this to where we can get a lot of our developers to just get hands on to where they can start learning it.

For the most part, it is the middleware team that uses this tool quite often, as well as other stuff in the low team where they use LoadRunner. In terms of troubleshooting, essentially it is the middleware team who is engaged with making sure that we have the Linux guest and that the applications are deployed, and they're all surfacey. That's what the middleware team is tasked with. In addition to that, the middleware team knows the product so well that they know how to troubleshoot it.

We go in there and we use it a lot, but one of our goals is to go back and try to engage some more of our application developer folks. Never too early, they could get more of a hands on.

Those guys are extremely helpful. If I were to rate them, I would probably give them a 9/10. The reason why I wouldn't give them a ten, because we had on a few occasions on a Friday, where we tried to reach individuals there in customer service and they weren't available. Even when we got them a message, the intake person said that we will contact somebody but they didn't. Then that Monday came around and they still didn't call us, so on Tuesday we called them and we said, "Hey, well, what happened? We had some issues and we let you guys know about it on Friday afternoon, and the on-call person wasn't called and nobody called us back." It wasn't a critical issue, but it was hot enough to where it would have been nice to get some guidance from Riverbed, but they went ahead and let us down, but that was only one occasion. That's why I wouldn't give them a ten, but they would definitely be at a very strong nine because during the day, like Monday through Friday during their core hours, they're always available, extremely responsive. Then they always get back to us and they also check in.

Let's say we're working on an issue. Periodically they'll reach back to us and say, "Hey, is everything okay? How is your testing going? Can we help you out with anything else?" They're very good on follow-up. They're very good on follow-up and they're extremely responsive.

We used to have another tool prior to Riverbed, which was CA Wiley but it was just not as robust. At one point we had found out that that tool had a memory leak in it which was causing our applications to go a little sideways. It took us three times of climbing the ladder to purchase Riverbed, and that was just convincing the executives that this was the way to go. Everybody likes it.

We kind of got hung on Riverbed. They came in and they did a demonstration for us, and we were so delighted with the demonstration that we kind of grabbed onto it, even though it was just a POC. We previously used another tool, we just didn't like it. It could not compare to Riverbed with the ease of use of the Riverbed suite. When we went to the Riverbed suite, we really latched onto it. It was very easy to set up. We really used it a lot, but then when the POC ended and it came to the point of making an executive decision on if we were going to purchase it or not, it went all the way to our executives and it was turned down. We weren't able to purchase it because apparently we had a long contract with CA, and the powers to be said, "Oh, continue to use FCIO because we won't have to pay for it."

The product is very useful to find problems in middleware for the application servers, especially agent instrumentation and management is user friendly. The main point for me is that there is no overhead on the application servers.

The Active Dashboards is valuable because we can show the results to all the related business units and then they can follow end2end system monitoring side to track it themselves.

The biggest problem of end-to-end APM monitoring is that service assurance is mapping a problem to the appropriate team. When the business units check their jobs, step by step, every single system seems to work properly but they need to check the big picture, which AppInternals does easily.

The admin dashboard could be easier as it takes a little bit of time to get used to it. Also, they need to better integrate with products and solutions from different vendors.

We have been using this for around two and a half years alongside AppResponse and BrowserMetrix.

10/10 as they have a quick response and are awesome.

We use different APM tools for addressing problems on the application side of the application servers. Every APM solution has a similar logic base, and through agents monitors traffic and renders the result. Our problem before using AppInternals was that the agent that worked on the application servers had a big overhead. This resulted in the owners of the applications not wanting to monitor with those agents. Therefore, we started to search for an APM solution for our applications. The starting point of our search was for a stable solution, and at the end of the POC process, we decided that the Riverbed AppInternals agent solved the overhead and stability problems best.

İt is really straightforward and can be completed with a couple of clicks and you don't need to perform a lot of deep diving.

The initial setup was performed by the vendor team, but updates and reinstalls have been done by in-house staff.

The licensing model is expensive compared to its competitors, but the service it gives to your business, and the data quality, means that it's worth it.

Every service assurance and monitoring team should try the product. It can probably improve your internal business relationships.

We can see problem trends before they affect customers and quickly drill down to where the issue is located. This eliminates support silos and allows various support teams to work together to quickly resolve the problem.

There are several valuable features, including:

• Seamless interaction between the various products;
• Ability to identify slow web pages, slow network/application transactions and drill down to offending packets, code, SQL statement, web service, network devices, or systems;
• Dashboard allows for "single pane of glass" monitoring of application and network performance; and
• Synthetic transactions, WMI and SNMP query capability.

• Product documentation and training need improvement. At one time, training was free and they supplemented training with road shows that helped users get more out of the product and, I think, helped Riverbed sell more products to the user community.
• Riverbed should also add the NetFlow module back to AppResponse.
• The recording mechanism for synthetic transactions could be improved as well.

We've had no issues with the stability.

RPM Dashboards can become unstable if too many items are added. This was supposedly fixed in Portal.

It's not as good as it was three years ago before meddling by an activist investor, privatization, and subsequent layoffs hurt support. They appear to be rebuilding but have a way to go.

We previously used something else and switched because Riverbed has faster problem resolution and the ability to see trends developing across the infrastructure. Many times what might have taken hours or days can be resolved in a few minutes.

It was mostly straightforward.

Negotiate with them.

Include some consulting hours in the purchase for a quicker startup.

Mean time to isolate a problem has dropped by 50-75% in each region since we've deployed the Riverbed SteelCentral suite. We're also able to isolate further at first look and select the correct resources to continue troubleshooting.

• Transaction Trace Warehouse (TTW) which records all transactions all the time, and lets us see performance down to the line of code.
• Browsermetrix, which is real-user monitoring via JS injection and linked back to TTW via cookie. It allows us to see the experience of every user hitting our sites and analyze performance by region, browser, etc.

Ability to trace transactions though non .Net/Java Layers has recently been added, but it's still dependent on a separate, traceable, unique ID.

Deployment and agent patch management is not managed centrally, resulting in a large level of effort to update.

We have had no issues with the stability.

In v8, the Dashboard and Browsermetrix products had issues with concurrent viewers. Riverbed supported us though modifications and incorporated a better design in v9.

Support from our sales team and internal product owners at Riverbed has been exceptional. The call-in help desk team is just OK.

We used and continue to use several traditional monitoring tools (SCOM), but we selected Riverbed for application performance monitoring, which grew to include network management and security use cases in the Riverbed ecosystem.

Initial setup was mostly very simple and has improved dramatically in v10. The slightly complex parts are around defining Java parameters and configuring Linux OS permissions.

We did POC deployments of Riverbed, AppDynamics, and CA APM. About 20 others were cut at RFI/RFP.

Your most immediate benefits will come from the ARX appliances, but the real rock star is TTW.

We spend less time determining the root cause of problems. It became easy to find out responsible groups for problems. We can discover performance problems on pre-production environments before delivering in production. This way, we can increase customer satisfaction.

• Acceptable overhead on application
• 100% transactions not sampling
• Clicks to root cause - fast & easy to diagnose and deep-dive
• Automated dynamic discovery of applications and mapping of transaction flows

We'd like to be able to find out performance problems on application class and methods.

• SQL performance problems: fast SQL execution but slow reading of the database. We solved these issues by increasing the fetch size in the code.
• Network problems: slow internet access by customers. There were CPU problems that couldn't be discovered by our Network Team.
• Backup problems: performance problems because of other systems which caused slow performance of this system.
• Database problems: performance problems between the database and storage but it told us it was an issue with the application.
• We had issues with the end-to-end view of unique customer: reviewing end- to-end customer problems to see where a problem is about customer, code, DB or network.

There have been no issues scaling it for our needs.

Support for issues when we need it is very good and professional in Turkey.

We used HP Diagnostics, Wily and Introscope Agent. All these couldn't help diagnose problems on our network or with applications. Also, they use sample data.

It's a bit complex to install and configure.

We implemented it through a vendor team. The vendor has great experience which helps to make it a success.

We did a PoC for AppDynamics, Dynatrace and Riverbed OPNET. Riverbed OPNET was more successful on technical support, better problem troubleshooting, and end-user monitoring.

If you want to save time for problem resolution, and to address problem on your environments you should try Riverbed OPNET.