mc ping reports "status=failed" despite exit code 0 and Error=0 against Ceph RGW

[JTRipper75]

Environment

mc (new):

mc version RELEASE.2025-08-13T08-35-41Z
Runtime: go1.24.6 linux/amd64

mc (working):

mc version RELEASE.2024-03-09T06-43-06Z
Runtime: go1.21.8 linux/amd64

Backend:

• Ceph RGW (S3-compatible)
• Same endpoint
• Same credentials
• Same alias configuration

Steps to reproduce

Configure an alias pointing to a Ceph RGW endpoint.

Run:

mc ping --count=1 <alias>
echo $?

Actual result

1: https://<endpoint>   status=failed  time=29.92ms

┌────────────────────────────────┬──────────┬──────────┬──────────┬───────┬───────┐
│ Endpoint                       │ Min      │ Avg      │ Max      │ Error │ Count │
│ https://<endpoint>             │ 29.92ms  │ 29.92ms  │ 29.92ms  │ 0     │ 1     │
└────────────────────────────────┴──────────┴──────────┴──────────┴───────┴───────┘

$ echo $?
0

The probe reports:

• status=failed
• Error=0
• exit code=0

These are contradictory.

Previous release behavior

With RELEASE.2024-03-09:

1: https://<endpoint> min=21ms max=21ms average=21ms errors=0 roundtrip=21ms

Same endpoint.
Same credentials.
Same backend.

Debug output

Both versions perform:

GET /minio/health/live

The backend is Ceph RGW.

The previous release handled the response as a successful probe.

The current release reports status=failed while still returning exit code 0 and Error=0.

Expected behavior

Either:

• report the probe as successful (consistent with exit code 0 and Error=0), or
• if the probe is considered failed, return a non-zero error count and non-zero exit status.

The current output appears internally inconsistent.