ARCHIVE/clio
1
0
Fork 0
mirror of https://github.com/XRPLF/clio.git synced 2026-07-23 23:20:24 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: Proxy support (#2490)

Browse Source 
Add client IP resolving support in case when there is a proxy in front
of Clio.
This commit is contained in:
Sergey Kuznetsov
2025-09-03 15:22:47 +01:00
committed by GitHub
parent 0a2930d861
commit 3a667f558c
39 changed files with 1042 additions and 125 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
docs/config-description.md
View File

@@ -329,6 +329,22 @@ This document provides a list of all available Clio configuration properties in
- **Constraints**: The minimum value is `1`. The maximum value is `4294967295`.
- **Description**: Maximum queue size for sending subscription data to clients. This queue buffers data when a client is slow to receive it, ensuring delivery once the client is ready.
### server.proxy.ips.[]
- **Required**: True
- **Type**: string
- **Default value**: None
- **Constraints**: None
- **Description**: List of proxy ip addresses. When Clio receives a request from proxy it will use `Forwarded` value (if any) as client ip. When this option is used together with `server.proxy.tokens` Clio will identify proxy by ip or by token.
### server.proxy.tokens.[]
- **Required**: True
- **Type**: string
- **Default value**: None
- **Constraints**: None
- **Description**: List of tokens in identifying request as a request from proxy. Token should be provided in `X-Proxy-Token` header, e.g. `X-Proxy-Token: <very_secret_token>'. When Clio receives a request from proxy it will use 'Forwarded` value (if any) to get client ip. When this option is used together with 'server.proxy.ips' Clio will identify proxy by ip or by token.
### prometheus.enabled
- **Required**: True

6
docs/examples/config/example-config.json
View File

@@ -76,7 +76,11 @@
"parallel_requests_limit": 10, // Optional parameter, used only if "processing_strategy" is "parallel". It limits the number of requests for one client connection processed in parallel. Infinite if not specified.
// Max number of responses to queue up before sent successfully. If a client's waiting queue is too long, the server will close the connection.
"ws_max_sending_queue_size": 1500,
"__ng_web_server": false // Use ng web server. This is a temporary setting which will be deleted after switching to ng web server
"__ng_web_server": false, // Use ng web server. This is a temporary setting which will be deleted after switching to ng web server
"proxy": {
"ips": [],
"tokens": []
}
},
// Time in seconds for graceful shutdown. Defaults to 10 seconds. Not fully implemented yet.
"graceful_period": 10.0,