> For the complete documentation index, see [llms.txt](https://docs.reviactyl.app/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.reviactyl.app/project/agent/additional-configuration.md). # Additional Configuration These are advanced configurations for Agent. You risk breaking Agent and making containers unusable if you misconfigure something. Proceed only if you know what each configuration value does. {% hint style="warning" %} You must apply all changes to your Agent `config.yml` file located at `/etc/reviactyl` and restart agent. Verify your config file using [Yaml Lint](http://www.yamllint.com/) should you receive errors related to YAML parsing. {% endhint %} ## Private Registries You can use these settings to authenticate against (private) docker registries when pulling images. ### Available Keys | Setting Key | Default Value | Notes | | ----------- | ------------- | ----------------- | | name | null | Registry address | | username | null | Registry username | | password | null | Registry password | ### Example of usage ```yaml docker: registries: registry.example.com: username: "registryusername" password: "registrypassword" ``` ## Custom Network Interfaces You can change the network interface that Agent uses for all containers by editing the network name; it is by default set to `reviactyl_nw`. For example, to enable Docker host mode change the network name to `host`. {% hint style="warning" %} Changing network mode to `host` grants Reviactyl direct access to all machine interfaces and Panel users can bind to any IP or Port even if it's not allocated to their container. You will lose all benefits of Docker network isolation. It is not recommended for public installations that are hosting other users' servers. {% endhint %} ### Example of usage ```yaml docker: network: name: host network_mode: host ``` After making changes, the following commands will stop the Agent, remove the Reviactyl network, and start the Agent again. Run at your own risk. `systemctl stop agent && docker network rm reviactyl_nw && systemctl start agent` ## Enabling Cloudflare proxy Cloudflare proxying of the Agent isn't beneficial since users will be connecting to the machine directly and bypassing any Cloudflare protection. As such, your Node machine IP will still be exposed. To enable Cloudflare proxy, you must change the Agent port to one of the Cloudflare HTTPS ports with caching enabled (more info [here](https://developers.cloudflare.com/fundamentals/get-started/reference/network-ports/)), such as 8443, because Cloudflare only supports HTTP on port 8080. Select your Node in the Admin Panel, and on the settings tab, change the port. Make sure that you set "Not Behind Proxy" when using Full SSL settings in Cloudflare. Then on Cloudflare dashboard, your FQDN must have an orange cloud enabled beside it. You are unable to proxy the SFTP port through Cloudflare unless you have their enterprise plan. ## Container PID Limit You can change the total number of processes that can be active in a container at any given moment by changing the `container_pid_limit` value. The default value is `512`. You can set it to `0` to disable the limit completely. However, this is *not* recommended as the limit prevents malicious overloading of the node. Restart agent and your game server to apply the new limit. ### Example of usage ```yaml docker: ... container_pid_limit: 512 ... ``` ## Throttles Limits You can use these settings to adjust or completely disable throttling. | Setting Key | Default Value | Notes | | ----------------------- | ------------- | ----------------------------------------------------------------------------------------------------------------------------------- | | enabled | true | Whether or not the throttler is enabled | | lines | 2000 | Total lines that can be output in a given line\_reset\_interval period | | maximum\_trigger\_count | 5 | Amount of times throttle limit can be triggered before the server will be stopped | | line\_reset\_interval | 100 | The amount of time after which the number of lines processed is reset to 0 | | decay\_interval | 10000 | Time in milliseconds that must pass without triggering throttle limit before trigger count is decremented | | stop\_grace\_period | 15 | Time that a server is allowed to be stopping for before it is terminated forcefully if it triggers output throttle | | write\_limit | 0 | Impose I/O write limit for backups to the disk, 0 = unlimited. Value greater than 0 throttles write speed to the set value in MiB/s | | download\_limit | 0 | Impose a Network I/O read limit for archives, 0 = unlimited. Value greater than 0 throttles read speed to the set value in MiB/s | ### Example of usage ```yaml throttles: enabled: true lines: 2000 maximum_trigger_count: 5 line_reset_interval: 100 decay_interval: 10000 stop_grace_period: 15 ``` ## Installer Limits Defines the limits on the installer containers that prevents a server's installation process from unintentionally consuming more resources than expected. This is used in conjunction with the server's defined limits. Whichever value is higher will take precedence in the install containers. | Setting Key | Default Value | Notes | | ----------- | ------------- | ----------------------------------------------------------------------------------------------------------- | | memory | 1024 | The maximum amount of memory install container can use unless server memory limit is higher than this value | | cpu | 100 | The maximum amount of cpu install container can use unless server cpu limit is higher than this value | ### Example of usage ```yaml installer_limits: memory: 1024 cpu: 100 ``` ## Other values More commonly discussed values. View all Agent config values and explanations in [these two files.](https://github.com/reviactyl/agent/tree/develop/config) | Setting Key | Default Value | Notes | | ------------------------------ | ------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------- | | debug | false | Force Agent to run in debug mode | | tmpfs\_size | 100 | The size of the /tmp directory in MB when mounted into a container | | websocket\_log\_count | 150 | The number of lines to display in the console | | detect\_clean\_exit\_as\_crash | true | Mark server as crashed if it's stopped without user interaction, e.g., not pressing stop button | | (crash detection) timeout | 60 | Timeout between server crashes that will not cause the server to be automatically restarted | | app\_name | "Reviactyl" | Changes the name of the daemon, shown in the panel's game console | | check\_permissions\_on\_boot | true | Check all file permissions on each boot. Disable this when you have a very large amount of files and the server startup is hanging on checking permissions | --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.reviactyl.app/project/agent/additional-configuration.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.