Log Completed Request: LogLevel off not work

[rvojcik]

Describe the bug
Default log_requests_level should be off by the documentation. By default Vault log all completed requests to stdout.
I tried to explicitly turn it off via config.json file "log_requests_level": "off",

My configuration

    "log_level": "error",
    "log_requests_level": "off",

Vault still log every completed request even success ones as errors

Nov 16 10:06:55 vault3 vault[6098]: 2023-11-16T10:06:55.118+0100 [ERROR] core: completed_request: start_time="2023-11-16T10:06:55+01:00" duration=1ms client_id=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx client_address=xxxxxxxxxxxxxxxxxxxx status_code=200 request_path=/v1/secrets/data/xxxxxxxxxxxxxxxxxxxx request_method=GET
Nov 16 10:06:55 vault3 vault[6098]: 2023-11-16T10:06:55.122+0100 [ERROR] core: completed_request: start_time="2023-11-16T10:06:55+01:00" duration=1ms client_id=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx client_address=xxxxxxxxxxxxxxxxxx status_code=200 request_path=/v1/certs/data/xxxxxxxxxxxxxxxxxxxxxxx request_method=GET

To Reproduce
Steps to reproduce the behavior:

1. Run vault server with disabled log_requests_level and log_level error
2. Run read data from vault
3. See stdout/stderr logs from vault server

Expected behavior
See only error (log_level: error) and no completed_request is logged (log_requests_level: off)

Environment:

• Vault Server Version (retrieve with vault status): 1.15.2
• Vault CLI Version (retrieve with vault version): 1.15.2
• Server Operating System/Architecture: Debian 10 (Buster), AMD64,

Vault server configuration file(s):
config.json

{
     "api_addr": "https://xxxx:8200",
    "cluster_addr": "https://xxx:8201",
    "default_lease_ttl": "6h",
    "disable_mlock": true,
    "listener": {
        "tcp": {
            "address": "0.0.0.0:8200",
            "tls_cert_file": "xxxxx",
            "tls_key_file": "xxxxx"
        }
    },
    "log_level": "error",
    "log_requests_level": "off",
    "storage": {
        "raft": {
            "node_id": "vault3",
            "path": "/x/raft_storage"
        }
    },
    "telemetry": {
        "disable_hostname": true,
        "dogstatsd_addr": "localhost:8125",
        "prometheus_retention_time": "30s"
    },
    "ui": true

}

[delamart]

Can confirm I can't disable request logs which in Kubernetes shows every probe call. Using vault:1.15.1-ubi

[adcsikos]

Same issue in my side as well! With vault verison:1.15.1-ubi. on openshift

[Dasio]

Probably fixed by #24056

[miagilepner]

Probably fixed by #24056
This is correct. This change will be in the next minor release, so I'm going to close this issue.