Acess Denied for S3 buckets is not reporting

[Roushan45]

Environment info

• NooBaa Version: noobaa-core-5.15.4-20240704.el9.x86_64
• Platform: RHEL9

Actual behavior

see the below o/p specially the bold one, dir permission for user s3user-1-dir is 770 and there is bucket associated to it bucket-new-1 for which I set the chmod to 000, no permission at all, then why there is no invalid bucket[ACCESS_DENIED] in health command output of /usr/local/noobaa-core/bin/node /usr/local/noobaa-core/src/cmd/health --all_bucket_details | jq | less
If i did same for s3user-1-dir then am getting the ACCESS_DENIED but for bucket there is no error for same scenario
[root@cluster4-41 ~]# ll /mnt/cesSharedRoot/
...
drwxrwx---. 3 10 wheel 4096 Jul 26 19:03 s3user-10-dir
drwxrwx---. 3 bin bin 4096 Jul 26 19:02 s3user-1-dir
drwxrwx---. 3 daemon daemon 4096 Jul 26 19:03 s3user-2-dir
drwxrwx---. 3 adm sys 4096 Jul 26 19:03 s3user-3-dir
...
[root@cluster4-41 ~]# ll /mnt/cesSharedRoot/s3user-1-dir
total 1
d---------. 2 bin bin 4096 Jul 26 19:02 bucket-new-1
[root@cluster4-41 ~]# /usr/local/noobaa-core/bin/node /usr/local/noobaa-core/src/cmd/health --all_bucket_details | jq | less
{
"service_name": "noobaa",
"status": "OK",
"memory": "267.6M",
"checks": {
"services": [
{
"name": "noobaa",
"service_status": "active",
"pid": "71215",
"error_type": "PERSISTENT"
}
],
"endpoint": {
"endpoint_state": {
"response": {
"response_code": "RUNNING",
"response_message": "Endpoint running successfuly."
},
"total_fork_count": 2,
"running_workers": [
2,
1
]
},
"error_type": "TEMPORARY"
},
"buckets_status": {
"invalid_buckets": [], <-- no invalid bucket
"valid_buckets": [
{
"name": "bucket-1",
"storage_path": "/mnt/cesSharedRoot/s3user-1-dir/bucket-new-1"
},
{
"name": "bucket-6",
"storage_path": "/mnt/cesSharedRoot/s3user-6-dir/bucket-new-6"
},
{
"name": "bucket-9",
"storage_path": "/mnt/cesSharedRoot/s3user-9-dir/bucket-new-9"
},

Expected behavior

1. bucket having no access should be reported in invalid buckets in health CLI

Steps to reproduce

1. manually changed the permission of the bucket folder to 000

More information - Screenshots / Logs / Other output

[romayalon]

thank you @Roushan45
we can definitely add a check for that, initially, we did only existence check, and I added also permissions check for account's new_buckets_path, so we need to add permissions check for that bucket's path.

[romayalon]

Hey @Roushan45, we merged a fix for this issue, available on master, hence closing the bug, can you please validate?
Fixing PR - #8488