Description
When I enable the AuditLogger feature in the config ([AuditLogger] Enabled = true), the node crashes immediately on startup with this error:
fatal error instantiating application: failed to register health check for service *audit.AuditLoggerService: misconfigured check "" for *audit.AuditLoggerService
It crashes right away, before the node even tries to connect to anything. It doesn't matter what ForwardToUrl is set to — I tested with both a working URL and a broken one, same crash either way. Turning off AuditLogger (or setting Enabled = false) lets the node start normally.
Basic Information
- Network: N/A — the node never gets far enough to connect to a chain, it crashes during startup
- Blockchain Client: N/A
- Go Version: 1.26.5 (from the build logs)
- Operating System: Ubuntu 24.04, running inside Docker
- Commit:
version=2.61.0@unset
- Hosting Provider: self-hosted
- Startup Command:
chainlink -c config.toml -s secrets.toml node start, also reproduced via docker run smartcontract/chainlink:2.61.0 node -c config.toml -s secrets.toml start
Environment Variables
Not using environment variables to configure the node — using config.toml / secrets.toml files instead.
Steps to Reproduce
- Add this to
config.toml:
[AuditLogger]
Enabled = true
ForwardToUrl = 'http://any-endpoint:9898/audit'
- Start the node normally.
- It crashes immediately with the error above.
Additional Information
While digging into it, it looks like the code that sets up the AuditLogger service passes it a logger object that was never given a name. When the node tries to register this service for health checks, it requires every service to have a name, and rejects it because the name is empty — which crashes the whole startup.
So basically, just turning this feature on (regardless of any other config) makes the node unable to start at all.
Description
When I enable the AuditLogger feature in the config (
[AuditLogger] Enabled = true), the node crashes immediately on startup with this error:It crashes right away, before the node even tries to connect to anything. It doesn't matter what
ForwardToUrlis set to — I tested with both a working URL and a broken one, same crash either way. Turning offAuditLogger(or settingEnabled = false) lets the node start normally.Basic Information
version=2.61.0@unsetchainlink -c config.toml -s secrets.toml node start, also reproduced viadocker run smartcontract/chainlink:2.61.0 node -c config.toml -s secrets.toml startEnvironment Variables
Not using environment variables to configure the node — using
config.toml/secrets.tomlfiles instead.Steps to Reproduce
config.toml:Additional Information
While digging into it, it looks like the code that sets up the AuditLogger service passes it a logger object that was never given a name. When the node tries to register this service for health checks, it requires every service to have a name, and rejects it because the name is empty — which crashes the whole startup.
So basically, just turning this feature on (regardless of any other config) makes the node unable to start at all.