Genesys Cloud - Main

 View Only

Sign Up

  • 1.  Architect Execution Data Usage

    Posted 15 days ago

    Hi everyone,

    I'm curious how others manage Architect Execution Data in production environments.

    How do you usually track the storage usage generated by Execution Data, ideally down to individual flows?

    Also, how do you decide which debugging level (Base, Notes, Verbose Notes, All) to use for your flows? Do you have a general standard, or do you adjust the level based on flow volume and troubleshooting needs?

    Interested to hear what approaches or best practices others are using.

    I usually go with Verbose Notes until Go-Live or end of hypercare, then take it down a notch to Base or Notes for BAU.


    #ArchitectandDesign
    #Reporting/Analytics
    #Routing(ACD/IVR)
    #System/PlatformAdministration

    ------------------------------
    Dominik
    Principal Professional Services Consultant
    ------------------------------


  • 2.  RE: Architect Execution Data Usage

    Posted 15 days ago

    Hi Dominik,

    We take a similar risk-based approach. Our designs separate PROD from TEST/DEV, use live-like testing, and build Flow Outcomes and Milestones into the flow design for ongoing visibility.

    We don't have a documented standard for Base, Notes, Verbose Notes or All, but using Verbose Notes through testing, go-live and hypercare, then moving to Notes or Base for BAU, fits the way we operate. For incidents, we'd temporarily increase the detail on the affected flow rather than leave verbose execution data enabled broadly in production.

    Per-flow storage visibility is the gap for us too. I haven't seen anything in our designs that measures or alerts on Architect Execution Data usage by individual flow, so I'd be interested to hear how others are handling that.



    ------------------------------
    Akhil
    ------------------------------



  • 3.  RE: Architect Execution Data Usage

    Posted 14 days ago

    I've never really give much thought to storage, it has never been an issue for us.

    If we can (if we are below the variable limit), we always use "All". Ultimately there is nothing more frustrating than trying to troubleshoot an execution "replay" but finding that variable values aren't logged and you can't see why a logic block failed.



    ------------------------------
    James
    Telecoms Specialist
    ------------------------------



  • 4.  RE: Architect Execution Data Usage

    Posted 14 days ago

    Hi Dominik !

    For me, Verbose Notes is very useful during testing, Go-Live and hypercare, because at that stage we normally need more visibility to understand what is happening inside the flow and troubleshoot issues faster.

    Once the flow is stable and moves to BAU, I prefer to reduce it to Base or Notes, especially for high-volume flows. I would only increase it again temporarily when we need to investigate a specific issue.

    For storage, I think this is also an important point. It would be very useful to have a simple way to understand which flows are generating more Execution Data and how much storage each one is consuming. This would help us make better decisions about the debugging level, especially in large production environments.

    So, in general, my approach is:

    Testing / Go-Live / Hypercare = Verbose Notes
    Stable production / BAU = Base or Notes
    Specific troubleshooting = Temporarily increase the level

    I think the key is finding the right balance between having enough information for troubleshooting and avoiding unnecessary data in production.



    ------------------------------
    Carlos
    Solution Consultant
    ------------------------------



  • 5.  RE: Architect Execution Data Usage

    Posted 13 days ago

    Hi Dominik

    For me, the Notes level is usually sufficient for standard troubleshooting.
     
    I prefer keeping things simple by using a tiered strategy:
     
    Global Baseline: I keep the organization default at Base. This protects high-volume IVRs from unnecessary logging noise while still capturing routing paths and critical errors.
     
    Flow-Level Overrides: I evaluate each flow's configuration individually based on its needs:
     
    None: I set high-volume flows with heavy interactions to None to minimize processing overhead, temporarily bumping them to Notes or Verbose Notes only when a specific issue needs investigation.
     
    Notes: Used for core business flows where tracking variable states is essential for diagnosing issues.
     
    Verbose Notes / All: Reserved strictly for active debugging on integration/Data Action flows or during the 1–2 week hypercare window post-release.
     
    All (Bot Interactions): I typically reserve the All level for scenarios where I need to analyze detailed message exchanges between a customer and a bot (such as AVA or Voice Bots) --> "What the Client Said --> What the Bot Received?"
     
    Security Check (Crucial): If you elevate to Verbose Notes or All, always ensure sensitive fields (tokens, PII, account numbers) are marked as Secure Variables in Architect so they aren't stored in plain text.
     
    A few extra tips & limits to keep in mind:
     
    1. Genesys Cloud only retains execution logs for 10 days.
    2. It only returns 200 execution instances, not necessarily the last 200. But you can change the query used to find the specific interaction you want to track.
    3. Flows with more than 350 variables will not display in the Execution Data view (replay journey).
    4. JSON, String, and Collection variable types have size limits; if a payload exceeds the threshold, the data won't render in the execution journey.
     
    For more details, you can check these resources:
     
    https://help.genesys.cloud/articles/historical-execution-data-overview/
    https://help.genesys.cloud/articles/manage-historical-execution-data/
    https://help.genesys.cloud/articles/genesys-cloud-fair-use-policy/
    I hope I've helped!!!!


    ------------------------------
    Cristian Rojas
    Interacciones Inteligentes S.A.
    ------------------------------



  • 6.  RE: Architect Execution Data Usage

    Posted 13 days ago

    We have also used all on our flows all the time since this te only mode that actually give you some value in troubleshooting. We have never had any problems with storage either and preformance. So to be able to troubleshoot I suggest use all.



    ------------------------------
    EysteinKylland
    DnB Bank ASA
    Systemutvikler Genesys Applikasjonsspesialist
    ------------------------------



  • 7.  RE: Architect Execution Data Usage

    Posted 13 days ago

    Thank you everyone for your feedback on this topic! It is very interesting to see some following best practice and tuning the debug levels to their needs and some leaving on Verbose or higher to have the flexibility of investigating any past interaction in detail.

    I did come across a section under  Menu > Account > Subscription > Storage where the accumulated Flow History storage for any given month is tracked. I was surprised at how little storage (less than 1GB) a month of Verbose Notes would generate in a fairly busy Org. Given that, I am now a bit less worried of having to set it to the correct level for a given project phase.



    ------------------------------
    Dominik
    Principal Professional Services Consultant
    ------------------------------



  • 8.  RE: Architect Execution Data Usage

    Posted 13 days ago

    Hi @Dominik Buehler,

    For me Notes level is enough for troubleshooting. My coleagues and our customer use it a lot, execution data is very helpful to see how a interaction behave and check conditions. 

    Some feedback from our customers about execution data, they called "architect log", they would like see some improvements like:

    • not show 200 interaction but the last 200 houndred intead for example
    • be able to invoke execution data directly from interactions view from a interaction

    Regards,

    SG



    ------------------------------
    Saugort Dario Garcia
    Arquitecto de soluciones
    e-Contact
    https://www.e-contact.cl/
    ------------------------------



  • 9.  RE: Architect Execution Data Usage

    Posted 12 days ago
    Edited by Jeremie SIMON 12 days ago

    Hi @Dominik Buehler,

    We use it temporarly at debugging level "All" to search and find errors in call flow like : Intermittent Participant Data Issue Between Architect Flows | Genesys Cloud - Main

    It's very useful

    Regards



    ------------------------------
    Jeremie
    IT
    ------------------------------



  • 10.  RE: Architect Execution Data Usage

    Posted 12 days ago

    Hi Dominik,

    I think your storage finding changes the equation a bit. If Verbose Notes generates relatively little Flow History storage, I would be more inclined to optimize for troubleshooting value rather than storage alone.

    My concern with dropping everything to Base is that intermittent issues are often discovered after the fact. Raising the level later will not give you the missing detail for the interaction that already failed.

    I would probably use Notes as the default, keep Verbose Notes / All for integration-heavy or business-critical flows, and reduce it only where volume or data sensitivity justifies it.

    Per-flow storage reporting would be the missing piece here. It would make this much easier to manage objectively.



    ------------------------------
    Cesar
    INDRA COLOMBIA
    ------------------------------