A request is counted every time your application downloads a configuration file from the ConfigCat CDN. The frequency of these downloads is totally under your control. Between downloads you can access your feature flags as many times as you like, it still counts as one.
Keep track of the number of requests your apps are making on the ConfigCat Dashboard.
Check out our architectural overview.
Current request/month limits
If you hit this limit, we will keep your application up and running. However, you can expect us contacting you on how we can meet your needs.
What is a request?
The ConfigCat SDKs - which you import into your applications - download your feature flags and settings in the form of a config.json from the ConfigCat servers and cache it locally. One config.json download counts as 1 request.
What is NOT a request?
Reading feature flag and setting values from the cache is not considered a request.
You can call
GetValue() as many times as you like.
These calls are always evaluated locally.
There are 3 different ways (polling modes) to control caching.
This animation explains the different polling modes:
Auto polling (default)
In auto polling mode, the ConfigCat SDK downloads the latest values automatically and stores them in the cache. This is done in every 60 seconds by default. You can set the polling interval to any number between 1 second and int max.
In lazy loading mode, the ConfigCat SDK downloads the latest setting values only if they are not present in the cache, or if the cache has expired. You can set the cache Time To Leave (TTL) to any number also.
Manual polling gives you full control over when the setting values are downloaded.
The ConfigCat SDK will not update them automatically.
You can (and should) update the cache manually, by calling a
forceRefresh() - this will download the latest values and update the cache.
Additionally, there are Webhooks available to get notified about changes instantly. Read more about Webhooks.
Requests explained through examples
Backend applications typically have a lower number of instances than frontend applications.
Example 1 - average polling with 4 instances
Let’s say you have an API for your frontend application and you have 4 instances of them behind a load balancer. All these 4 instances use ConfigCat SDK in auto polling mode with a 1 minute polling interval.
4 (instances) 43,800 (minutes in a month)* = 175,200 requests / month
Example 2 - quick polling with 10 instances
If you want your system to react faster after changing a feature flag in ConfigCat, you can decrease the default polling interval down to 1 seconds. In this case we are calculating with 10 running instances.
10 (instances) 2,592,000 (seconds in a month)* = 25,920,000 requests / month
Typically, you have a high number of frontend/mobile/desktop application instances. Their number is determined by your user count.
Example 3 - a web app with 15,000 users
Web apps run in the browser, so for each user there will be a different ConfigCat SDK instance running. In this example, we have 15,000 active users who usually spend 5 hours on your web application per month. The ConfigCat SDK is set to Auto polling mode with 60 seconds polling interval.
18,000 (5 hours in seconds) / 60 polling interval = 300 requests/user/month
15,000 (users) 300 (requests/user/month)* = 4,500,000 requests / month
Example 4 - a mobile app
Having a mobile app which runs on the devices as a background process. The ConfigCat SDK is set to Auto polling mode with 1 hour polling interval.
Your application runs on 5,000 devices.
5,000 (devices) 730 (hours in a month)* = 3,650,000 requests / month