Hi,
I commented a similar issue a couple of months ago:
https://lists.ceph.io/hyperkitty/list/ceph-users@xxxxxxx/message/IQX2VXA6QQQPEZQ7GU3QY2WPHAIVPIUN/
Can you check if that applies to your cluster?
Zitat von Magnus Larsen <magnusfynbo@xxxxxxxxxxx>:
Hi Ceph-users!
Ceph version: ceph version 17.2.6
(d7ff0d10654d2280e08f1ab989c7cdf3064446a5) quincy (stable)
Using cephadm to orchestrate the Ceph cluster
I’m running into https://tracker.ceph.com/issues/59189, which is
fixed in next version—quincy 17.2.7—via
https://github.com/ceph/ceph/pull/50906
But I am unable to upgrade to the fixed version because of that bug
When I try to upgrade (using “ceph orch upgrade start –image
internal_mirror/ceph:v17.2.7”), we see the same error message:
executing _write_files((['dkcphhpcadmin01', 'dkcphhpcmgt028',
'dkcphhpcmgt029', 'dkcphhpcmgt031', 'dkcphhpcosd033',
'dkcphhpcosd034', 'dkcphhpcosd035', 'dkcphhpcosd036',
'dkcphhpcosd037', 'dkcphhpcosd038', 'dkcphhpcosd039',
'dkcphhpcosd040', 'dkcphhpcosd041', 'dkcphhpcosd042',
'dkcphhpcosd043', 'dkcphhpcosd044'],)) failed. Traceback (most
recent call last): File "/usr/share/ceph/mgr/cephadm/ssh.py", line
240, in _write_remote_file conn = await
self._remote_connection(host, addr) File
"/lib/python3.6/site-packages/asyncssh/scp.py", line 922, in scp
await source.run(srcpath) File
"/lib/python3.6/site-packages/asyncssh/scp.py", line 458, in run
self.handle_error(exc) File
"/lib/python3.6/site-packages/asyncssh/scp.py", line 307, in
handle_error raise exc from None File
"/lib/python3.6/site-packages/asyncssh/scp.py", line 456, in run
await self._send_files(path, b'') File
"/lib/python3.6/site-packages/asyncssh/scp.py", line 438, in
_send_files self.handle_error(exc) File
"/lib/python3.6/site-packages/asyncssh/scp.py", line 307, in
handle_error raise exc from None File
"/lib/python3.6/site-packages/asyncssh/scp.py", line 434, in
_send_files await self._send_file(srcpath, dstpath, attrs) File
"/lib/python3.6/site-packages/asyncssh/scp.py", line 365, in
_send_file await self._make_cd_request(b'C', attrs, size, srcpath)
File "/lib/python3.6/site-packages/asyncssh/scp.py", line 343, in
_make_cd_request self._fs.basename(path)) File
"/lib/python3.6/site-packages/asyncssh/scp.py", line 224, in
make_request raise exc asyncssh.sftp.SFTPFailure: scp:
/tmp/etc/ceph/ceph.conf.new: Permission denied During handling of
the above exception, another exception occurred: Traceback (most
recent call last): File "/usr/share/ceph/mgr/cephadm/utils.py", line
79, in do_work return f(*arg) File
"/usr/share/ceph/mgr/cephadm/serve.py", line 1088, in _write_files
self._write_client_files(client_files, host) File
"/usr/share/ceph/mgr/cephadm/serve.py", line 1107, in
_write_client_files self.mgr.ssh.write_remote_file(host, path,
content, mode, uid, gid) File "/usr/share/ceph/mgr/cephadm/ssh.py",
line 261, in write_remote_file
self.mgr.wait_async(self._write_remote_file( File
"/usr/share/ceph/mgr/cephadm/module.py", line 615, in wait_async
return self.event_loop.get_result(coro) File
"/usr/share/ceph/mgr/cephadm/ssh.py", line 56, in get_result return
asyncio.run_coroutine_threadsafe(coro, self._loop).result() File
"/lib64/python3.6/concurrent/futures/_base.py", line 432, in result
return self.__get_result() File
"/lib64/python3.6/concurrent/futures/_base.py", line 384, in
__get_result raise self._exception File
"/usr/share/ceph/mgr/cephadm/ssh.py", line 249, in
_write_remote_file logger.exception(msg)
orchestrator._interface.OrchestratorError: Unable to write
dkcphhpcmgt028:/etc/ceph/ceph.conf: scp:
/tmp/etc/ceph/ceph.conf.new: Permission denied
We were thinking about removing the keyring from the Ceph
orchestrator
(https://docs.ceph.com/en/latest/cephadm/operations/#putting-a-keyring-under-management),
which would then make Ceph not try to copy over a new ceph.conf,
alleviating the problem
(https://docs.ceph.com/en/latest/cephadm/operations/#client-keyrings-and-configs),
but in doing so, Ceph will kindly remove the key from all nodes
(https://docs.ceph.com/en/latest/cephadm/operations/#disabling-management-of-a-keyring-file)
leaving us without the admin keyring. So that doesn’t sound like a
path we want to take :S
Does anybody know how to get around this issue, so I can get to
version where the issue fixed for good?
Thanks,
Magnus
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
