oms4suse / sesceph Goto Github PK

On slower clusters the timeout value of 5 is too low.

/usr/bin/ceph --connect-timeout 50 --keyring /var/lib/ceph/bootstrap-osd/ceph.keyring --name client.bootstrap-osd -f json-pretty status
works,
while
/usr/bin/ceph --connect-timeout 5 --keyring /var/lib/ceph/bootstrap-osd/ceph.keyring --name client.bootstrap-osd -f json-pretty status

doesn't due to timeouts.

For debugging issues with sesceph it would be nice to have more debugging output of what it does when calling salt with a higher debug level.

When f.e. calling

salt-call -l all  sesceph.keyring_admin_create 

no debugging output of the called
execute_local_command
can be found in the debug output of salt.

execute_local_command at https://github.com/oms4suse/sesceph/blob/master/_modules/sesceph/utils.py#L22
calls
salt['cmd.run_all']... with a loglevel
and therefore some output of salt['cmd.run_all'] should be found in debugging output of salt with a loglevel equal or higher than trace.

Even changing the loglevel in
https://github.com/oms4suse/sesceph/blob/master/_modules/sesceph/utils.py#L25
doesn't help.

Compared to the lvm module when calling

salt-call -l debug lvm.lvcreate salttest vgceph size=1G

some lines of salt['cmd.run'] ... can be found in debugging output

[DEBUG   ] LazyLoaded cmd.run
[INFO    ] Executing command ['/usr/sbin/lvcreate', '-n', 'salttest', 'vgceph', '-L', '1G'] in directory '/root'
[ERROR   ] Command ['/usr/sbin/lvcreate', '-n', 'salttest', 'vgceph', '-L', '1G'] failed with return code: 5
[ERROR   ] output:   Logical volume "salttest" already exists in volume group "vgceph"
[INFO    ] Executing command ['lvdisplay', '-c', '/dev/vgceph/salttest'] in directory '/root'
[DEBUG   ] stdout:   /dev/vgceph/salttest:vgceph:3:1:-1:0:2097152:256:-1:0:-1:251:6

(see also https://github.com/saltstack/salt/blob/develop/salt/modules/linux_lvm.py#L341)