You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi all,
We have recently started with the Experiment SDK. I was expecting pollOnStart: true to instantiate a poll mechanism that would fetch the user flags and variants on a regular basis. Inspecting the network requests and the source code, however, we found that the flag definitions are fetched every 60 seconds, but the values for the current users are not re-evaluated automatically as part of this process.
I wonder, what is the purpose of fetching the flags so often without re-evaluating the flag values? I expect the value to change more often, as this depends on user properties/events that might affect the target audience of a variant.
The text was updated successfully, but these errors were encountered:
You're right, pollOnStart only polls for flag configurations, it does not periodically fetch remote variants.
pollOnStart is mostly useful for local evaluation, which is a less common use case on the client-side. Unless you're using local evaluation I would recommend setting pollOnStart: false. If you want to re-fetch remote evaluation variants at an interval you would need to implement that manually in your app using setInterval or something similar.
Thanks for clarifying @bgiori . Do you have a recommendation about how often it would be good to poll for this flag? Would it be a problem to fetch the flag values every minute, as pollOnStart does by default?
Hi all,
We have recently started with the Experiment SDK. I was expecting
pollOnStart: true
to instantiate a poll mechanism that would fetch the user flags and variants on a regular basis. Inspecting the network requests and the source code, however, we found that the flag definitions are fetched every 60 seconds, but the values for the current users are not re-evaluated automatically as part of this process.I wonder, what is the purpose of fetching the flags so often without re-evaluating the flag values? I expect the value to change more often, as this depends on user properties/events that might affect the target audience of a variant.
The text was updated successfully, but these errors were encountered: