How to right size when you are over LaunchDarkly system resource count limits – LaunchDarkly

Overview

To protect the core infrastructure of LaunchDarkly and provide a predictable and reliable experience for our customers, we must limit the number of resources that can be created. We intend to set limits reasonably high enough so that most customers will never reach them. In some rare cases, larger customers may reach the limit and need to undergo an exception process to obtain a temporary higher limit. Current customers already over the limit will be given an exception to the default limits.

Solution

The following are best practices:

If you are close to or over the limit, review your resources and delete those no longer in use.
Develop ongoing resource hygiene policies.
If you are creating resources for testing:
- Delete any resources that were created during the tests.
- Consider more sustainable strategies: Testing code that uses feature flags.

If you continue to encounter resource limits even after applying the solution below, submit a support ticket. We'll work with you to either raise the limit or optimize your instance to utilize the resource in question better.

Review the table below for how to resolve the specific issue.

Resource Type	Solution	Hard Limit*
`Flag`	Review your flag hygiene policies. As best practice, it is recommended that you clean up and remove flags that are no longer needed.	5,000 per project
`Project`	Typically, if you are dynamically creating projects, this would be for testing purposes. In this case, make sure you are also proactively deleting them, at the very least. However, there are better-recommended strategies for testing with LaunchDarkly, see Testing code that uses feature flags.	300 per account
`Environment`	Typically, if you are dynamically creating environments, this would be for testing purposes. In this case, make sure you are also proactively deleting them at the very least. However, there are better-recommended strategies for testing with LaunchDarkly, see Testing code that uses feature flags. Some customers may encounter this if they have an environment per developer.	300 per project
`Access Tokens`	If you are dynamically creating access tokens, you should also proactively delete those access tokens.	200 per account
`Custom Role`	You may naturally encounter this if you need granular control over several teams and projects. If you don’t believe that all custom roles are needed, you can look for ways to optimize by combining policies, using wildcards (`*`) or by using `notResources`/ `notActions`. Review your resources and delete those that are no longer in use.	1,000 per account
`Team`	Review your resources and delete those that are no longer in use.	100 per account
`Team Custom Roles`	Review your resources and delete those that are no longer in use.	50 per team
`Team Members`	Review members in the team and remove members that are no longer with that team.	1,000 per team
`Relay Proxy Configuration`	Review your resources and delete those that are no longer in use.	50 per account
`Integrations Webhooks`	Review your resources and delete those that are no longer in use.	100 per account
`Flag Triggers`	Review your resources and delete those that are no longer in use.	20 per flag
`Segments`	Typically, multiple segments are used to individually target more than 10k users. Instead of using multiple segments, we recommend using Big Segments.	4,000 per environment
`Members`	Determined by contract/plan subscription

*Hard limits as of January 2025

Overview

Solution

Resource Type

Solution

Hard Limit*

Related articles