This is on main, but it's similar on reef/quincy. It comes from https://github.com/ceph/ceph/blob/main/src/pybind/mgr/cephadm/serve.py#L397 <https://github.com/ceph/ceph/blob/main/src/pybind/mgr/cephadm/serve.py#L397> which is part of a function that does the general refresh of daemons, devices, networks, etc. https://github.com/ceph/ceph/blob/main/src/pybind/mgr/cephadm/serve.py#L247 which is called from the main background service loop at https://github.com/ceph/ceph/blob/main/src/pybind/mgr/cephadm/serve.py#L84 On Thu, May 30, 2024 at 8:28 AM Yonatan Zaken <yzaken@redhat.com> wrote:

Hi, For a given ceph cluster I have deployed, I have a HEALTH_WARN. ceph health detail shows:

HEALTH_WARN failed to probe daemons or [WRN] CEPHADM_REFRESH_FAILED: failed to probe daemons or host HOST-1 `cephadm ceph-volume` failed: Cannot decode JSON

I have noticed that in cephadm.log there is a periodic print of the following command:

cephadm ['--image', '172.32.0.2:8787/ceph/daemon:reef-rockylinux-8-x86_64', '--no-cgroups-split', '--timeout', '895', 'ceph-volume', '--fsid', 'ca4a8ffc-2431-4cf8-9475-7156acdb8556', '--', 'inventory', '--format=json-pretty', '--filter-for-batch']

My aim is to figure out where, in the ceph source code, the above is executed, So far I noticed it triggers the method: _run_cephadm_json from : "/usr/share/ceph/mgr/cephadm/serve.py"

What I couldn't figure out is what triggers the method _run_cephadm_json?

BR, Yonatan _______________________________________________ Dev mailing list -- dev@ceph.io To unsubscribe send an email to dev-leave@ceph.io