Genesys Cloud - Developer Community!

I have created a custom data action and set it's timeout to 60s. I expect the request to take anywhere from 30-60s. When I go to test this in the data action configuration, it consistently returns a 504 gateway timeout after about 20s. When I inspect the request on our server, it received and processed the request as expected. This leads me to believe the genesys proxy is aborting the request sooner than configured. I have tried overriding this value in a call flow and get the same 20s response timeout. I know this because I logged the result to Participant Data on the interaction.

I have created a custom data action and set it's timeout to 60s. I expect the request to take anywhere from 30-60s. When I go to test this in the data action configuration, it consistently returns a 504 gateway timeout after about 20s. When I inspect the request on our server, it received and processed the request as expected. This leads me to believe the genesys proxy is aborting the request sooner than configured. I have tried overriding this value in a call flow and get the same 20s response timeout. I know this because I logged the result to Participant Data on the interaction.

the 20 second timeout is being enforced at the public API layer, and is not configurable.  The test execution will actually timeout at 19 seconds so that the timeout response can be returned within the Public API 20 second timeout window.  All of the relevant timeframes are documented here: https://help.mypurecloud.com/faqs/how-many-seconds-before-a-data-action-times-out/

Note that a 30 second operation/response is not ideal, as it's very likely that will cause the org to start to encounter the Data Action concurrent execution limit.  A much better paradigm to look at is a "jobs style" API, sending your query to a jobs endpoint that will queue and marshal the long running operation and return a job ID as the response.  That job ID can then  be utilized with a second data action that can query for the status of the job and return the results.  This allows both endpoints to be very responsive (returning their responses in under a second) while still allowing the long running operation to process.  The Genesys Cloud Analytics Jobs endpoints are a good example of this pattern.

I have created a custom data action and set it's timeout to 60s. I expect the request to take anywhere from 30-60s. When I go to test this in the data action configuration, it consistently returns a 504 gateway timeout after about 20s. When I inspect the request on our server, it received and processed the request as expected. This leads me to believe the genesys proxy is aborting the request sooner than configured. I have tried overriding this value in a call flow and get the same 20s response timeout. I know this because I logged the result to Participant Data on the interaction.

the 20 second timeout is being enforced at the public API layer, and is not configurable.  The test execution will actually timeout at 19 seconds so that the timeout response can be returned within the Public API 20 second timeout window.  All of the relevant timeframes are documented here: https://help.mypurecloud.com/faqs/how-many-seconds-before-a-data-action-times-out/

Note that a 30 second operation/response is not ideal, as it's very likely that will cause the org to start to encounter the Data Action concurrent execution limit.  A much better paradigm to look at is a "jobs style" API, sending your query to a jobs endpoint that will queue and marshal the long running operation and return a job ID as the response.  That job ID can then  be utilized with a second data action that can query for the status of the job and return the results.  This allows both endpoints to be very responsive (returning their responses in under a second) while still allowing the long running operation to process.  The Genesys Cloud Analytics Jobs endpoints are a good example of this pattern.

I have created a custom data action and set it's timeout to 60s. I expect the request to take anywhere from 30-60s. When I go to test this in the data action configuration, it consistently returns a 504 gateway timeout after about 20s. When I inspect the request on our server, it received and processed the request as expected. This leads me to believe the genesys proxy is aborting the request sooner than configured. I have tried overriding this value in a call flow and get the same 20s response timeout. I know this because I logged the result to Participant Data on the interaction.