Em sáb., 4 de jul. de 2020 às 11:27, Dave Hall <kdhall@binghamton.edu> escreveu:

Rodrigo,

I tried to send this to the list last night, but it looks like it didn't go through. I had a problem very much like this when I was first setting up my cluster. It turned out to be a missing systemd unit file. I would suggest that you check to see that you have an instance of ceph-volume@.service and an instance of ceph-osd@.service running for each OSD.

I manually started my ceph-volume@ instances and then I managed to successfully restart my ceph-osd@ instances. My OSDs are back. Does anybody knows how is responsible for calling the ceph-volume@ instances on boot? Regards, Rodrigo