Hi,

Couple of things that may help;

First doulbe check that your module is where it's supposed to be. For example, if you're seeing ENOENT, that seems to imply that the module is not in the /usr/share/ceph/mgr/ directory

e.g.

[root@ceph ~]# ceph mgr module enable thinair
Error ENOENT: all mgr daemons do not support module 'thinair', pass --force to force enablement

However, IIRC ENOENT is also returned if the module has been marked bad due to a prior failure. In this case disable and enable the module with --force - as the message states.

On Mon, Jan 13, 2020 at 3:08 AM Khoury, Rima (Nokia - IL/Kfar Sava) <rima.khoury@nokia.com> wrote:

Thank you! The logs did help but I had to add log_to_file (as true) in the ceph.conf in order to see it.
The plugin was successfully installed and did run as expected and could produce output as desired and implemented in the code. However, I tried to write a new function, but unfortunately it produced an error, but when I restored the working code I could not enable the plugin again. Now it says: " Error ENOENT: all mgr daemons do not support module 'newplugin', pass --force to force enablement".
I tried restarting the ceph-mgr, and made sure to do all the steps I did before, but seems that it's still stuck on this error.
Do you have any ideas of what might help? Could this be a bug in ceph?
BTW I am using nautilus

-----Original Message-----
From: Willem Jan Withagen <wjw@digiware.nl>
Sent: Thursday, January 9, 2020 1:41 PM
To: Juan Miguel Olmo Martinez <jolmomar@redhat.com>; Khoury, Rima (Nokia - IL/Kfar Sava) <rima.khoury@nokia.com>
Cc: dev@ceph.io
Subject: Re: New Ceph Plugin - assistance

On 9-1-2020 09:01, Juan Miguel Olmo Martinez wrote:
> Hi,
>
> Depending of the Ceph version you can have the files in different
> places, but in any case is easy to search the file "module.py" in any
> of your manager hosts, the results will show you the location of the
> manager plugins.
> In Nautilus they are in "/usr/share/ceph/mgr/"
>
> You can see the active manager using the command "ceph -s"
>
> A good source of information to see what is happening is the manager
> log file (in var/log/ceph) in the manager host,
>
> You can change the log level to debug using:
> ceph daemon <your-mgr> config set debug_mgr 20/5 restore the log
> level:
> ceph daemon <your-mgr> config set debug_mgr 1/5

I had a lot of trouble getting plugins to work, since they needed py-libs that I did not (yet) have.
Python tracebacks are logged in the ceph-mgr log file.
And there is was usually quite fast clear what the problem was.

Big chance that ceph-mgr is not loading you plugin due to loading errors but the traceback in the log should give you more info.

--WjW

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