Recommended Third Party Monitoring Tools for Instrumenting Flows

[ITTeam]

Hello Alteryx Server forums, 

We're a heavy Alteryx shop and recently we've been looking into setting up some sort of instrumentation for our Workflows to be able to better observe and monitor the performance and visibility of who is running what flow, how each flow is running, etc. We have DataDog for Server/Systems monitoring, however Application Performance Monitoring (APM) does not appear to be possible with Alteryx being a closed source, C++ application. Is there any other organization or person on here who has a recommendation for a third-party APM solution that could potentially work with Alteryx? Our end goal is being able to detect bottlenecks in performance beyond the typical server metrics (CPU, Memory, Bandwidth, Ops, Throughput, etc.) to drill down into flows that could be better written or to ultimately be more efficient. 

Appreciate any thoughts on this topic, thank you. 

Jonathan G
IT Team
TargetSmart Communications 

[TheCoffeeDude]

I haven't run into anything. However, you could query the Alteryx Gallery API for info and drop it into a database or a bunch of CSV files then have something like DataDog query it. I hear that there'll be a new Alteryx Gallery API that would give us more info about things on the server. There's no firm answer on when that'll drop.

[ITTeam]

Hi TheCoffeeDude,

Appreciate the reply -- that honestly sounds awful, as mentioned we have DataDog, but having to dump API data into .csvs then have it read/integrate sounds far from ideal. We have some experience in using the agent on other .NET and Java apps, just are stymied by Alteryx's seeming lack of visibility into how workflow performance can be measured or quantified. We're looking to detect bottlenecks, slow processes, errors in workflow configuration, etc. beyond typical server metrics that an Application Performance Monitoring (APM) solution can provide. Doesn't appear that this is possible with Alteryx -- open to other ideas, though. Thanks. 

[hroderick]

How about passing %stats% to capture as parms to a before run and/or after run event command line that writes them to a csv.

I haven't tried the %stats% that are used in email events but it might be worth a try

[raychase]

I've written several custom macros, leveraging the MongoDB Input connector, to query relevant data from the Alteryx Server repository and store it in a separate SQL environment.  From there, I've created various forms of analytical reporting, focused on run times, queue times, excessive scheduling, etc.

I've also scripted a job that downloads all workflows from the Alteryx Server and parses through their .xml content to identify poor development practices (ie. SELECT * queries or excessive flat-file input sources).

[ITTeam]

Hroderick and raychase -- thank you both, if there's a formal document or guide that you can provide to be able to set up your solutions properly, we're open to investigating it.

hroderick -- What sort of information does %stats% capture? Have you used this to identify bottlenecks in workflows before?
raychase -- If these are posted in the Public Gallery or elsewhere such as in a GitHub repo, or that script that parses all workflows for poor dev practices, that sounds pretty awesome and we'd definitely check this out. 

[hroderick]

This discusses what I meant by a %stat% variables used in an email event,   https://community.alteryx.com/t5/Alteryx-Designer-Discussions/Email-Events-Wildcard-Documentation/td...