Revisiting AppStack

After a long tenure working on the OrionRegistered Platform, I’ve recently shifted my responsibilities to fully focus on Server & Application Monitor (SAM). Features designed on the platform and in SAM have eye-opening similarities due to deep integration between SAM, Virtualization Manager (VMAN), Web Performance Monitor (WPM), and other heavy hitters in our systems portfolio. The same tenets of componentization and shareability demanded by the Orion Platform exist in AppStackTm the way they do for PerfStackTm or the newest generation of Orion maps.

In honor of this revelation and how far our integration story has come since the first introduction of AppStack in 2014, I’d like to revisit this milestone feature and show those new to the SolarWinds systems portfolio the power of what we provide. For those who enjoy nostalgia, revisit the first AppStack post here https://thwack.solarwinds.com/community/solarwinds-community/product-blog/blog/2014/11/03/appstack. Personally, I was taken aback by the amount of change that’s occurred in the UI itself.

Welcome to 2014, amirite? (I stole this screenshot from Jeremy's original 2014 post.)
appstack.png

Fast forward to 2019, the look and feel is quite different. Navigate to AppStack through the menu bar, or enjoy the contextual AppStack widget on the details page for an entity.

appstack1.png

appstack2.png

For those who land on the full AppStack view today, you'll notice we have new entities appearing in the stack with the inclusion of container monitoring.

pastedImage_6.png

When we as Product Managers introduce the capability to monitor new entities such as containers, we must first ask if it deserves a place in the AppStack. For containers, this is certainly true, due to their ephemeral nature and clear distinction as a generic entity type. The same can be said for the improvements to Cisco UCS monitoring, where SAM added chassis, blade, and rack server statistics into the AppStack view. However, in the case of VMware vSAN entities, you'll notice their inclusion into AppStack in a subtle approach aligned with customer expectations for hyperconverged infrastructure.

appstack3.png

In 2019.2 versions of the platform and later, the spotlight workflow is still an effective tool to quickly analyze where the problem might lie along your infrastructure stack.

2019.2 appstack spotlight.gif

The subtle difference lies in the changes to node status in the Orion Platform 2019.2 release. With simplified status calculations, and clear contributors detailed in the popovers, it's easier than ever to navigate to where you need to drill in for detailed troubleshooting. 

appstack4.png

With additional changes from VMAN 8.4 to add virtual entities as status contributors and the ability to control the status contributors via the Node status contributors page, the AppStack solution becomes even more powerful. Through continued improvement and integration throughout the Orion Platform and the system portfolio, AppStack in 2019 has aged well and can help you navigate the intricacies and quirks of your environment.

appstack5.png

Supplementing AppStack capabilities, through the addition of new Orion Maps and PerfStack, you now have a full toolset available to visualize your environment, narrow down the problem, and then troubleshoot the problem in-depth in real time.

Now that we've walked through how AppStack has grown over the years, I'd love to hear from you, both new and familiar to AppStack. What was your introduction to AppStack? Was it back in 2014 or the newer versions available today? What would you like to see improved in the future and what would you like to see preserved to keep the heart of AppStack beating strong for the next generation of Systems Management product releases? Put your feature request into Server & Application Monitor Feature Requests  for tracking and community input.

Anonymous
  • Since SCM’s release, I’ve always thought it was a great idea to monitor the directory where you store your scripts to make sure they didn’t vary and to validate changes over time, even going in and reverting them in case there was a change without approval. However, that part was available in the initial release of SCM. Using PowerShell with SCM, you can monitor your C:\Scripts\*.ps1 files and get notified when any deviate from their baselines.

    MyCCPay

  • Hi ,

    When I pass appstack an object it gives me the stack relating to that object. If I put it on a filtered summary page (Passing it 2+ objects) I get only the objects passed in, not the stacks that those objects lay within.

    This is understandable, though not the behaviour I would expect. I've build out a couple custom appstack views for this purpose and it was quite difficult. A behaviour toggle or something like that would be super appreciated. (or some simmilar method!)

    Or am I misunderstanding the use of appstack in these scenarios?

  • jfaldmo  wrote:

    I'm not sure what visually distinct icons mean. To me the purpose of the AppStack is to see where the problems are quickly. When there is so much red and yellow icons in the full AppStack view you really can't get to the root of a problem quickly. Maybe I'm doing it wrong and want to use AppStack for triage when an issue is occurring instead of using it just as one pane of glass view to jump to where any unspecified (I don't know what I am looking for) issue is.

    Interesting, what would you say is the root cause of so much red/yellow on your AppStack? has Orion Platform 2019.2 - Enhanced Node Status  helped with the intelligence of status?

  • I'm not sure what visually distinct icons mean. To me the purpose of the AppStack is to see where the problems are quickly. When there is so much red and yellow icons in the full AppStack view you really can't get to the root of a problem quickly. Maybe I'm doing it wrong and want to use AppStack for triage when an issue is occurring instead of using it just as one pane of glass view to jump to where any unspecified (I don't know what I am looking for) issue is.

  • I have the same issue/concern as you, shuth​, about seeing "the volumes/applications from BOTH servers." I put this idea in to make a separate category for applications that are from a different node.