Re: pg_num not being set to ceph.conf default when creating pool via python librados

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, 26 Jan 2015 11:16:34 -0800 Gregory Farnum wrote:

> Hmm, I'm looking at the actual code here and I'm wrong. Those values
> should be used whenever you create a pool via the API, and it doesn't
> look like anything external to the monitor can change that.
> 
> So you probably set these values in the [osd] section rather than the
> [mon] or [global] section, at a guess. (The monitor is the one who
> manages this data and so they need to have the values.)
>

If you look at his conf he clearly had this defined in the global section.

Given this:
http://tracker.ceph.com/issues/10012

and the various reports of these pg and pgp defaults NOT being picked up
during the initial pool creation I really think somebody needs to go over
the config parser with  fine comb.

>From the looks of it we have values that aren't picked up outside of
global and others that aren't outside their specific section.

Aside from fixing this, logging during startup (normal levels, no debug)
what the respective daemon did set from the config (and what it didn't
like) would go a long way.

Christian

> -Greg
> 
> On Mon, Jan 26, 2015 at 11:03 AM, Jason Anderson
> <Jason.Anderson@xxxxxxxxxxxxxxxx> wrote:
> > Hello Gregory,
> >
> >
> >
> > Thanks for the quick response. Does this mean that the rados python
> > library is out of date? create_pool in rados.py
> > (https://github.com/ceph/ceph/blob/master/src/pybind/rados.py#L535)
> > only requires a pool_name….doesn’t even offer pg_num as an optional
> > argument.
> >
> >
> >
> > Thank you,
> >
> > -Jason
> >
> >
> >
> > From: Gregory Farnum [mailto:greg@xxxxxxxxxxx]
> > Sent: Monday, January 26, 2015 10:09 AM
> > To: Jason Anderson; ceph-users@xxxxxxxxxxxxxx
> > Subject: Re:  pg_num not being set to ceph.conf default
> > when creating pool via python librados
> >
> >
> >
> > Just from memory, I think these values are only used for the initial
> > pool creations when the cluster is first set up.
> >
> > We have been moving for a while to making users specify pg_num
> > explicitly on every pool create, and you should do so. :)
> > -Greg
> >
> > On Mon, Jan 26, 2015 at 7:38 AM Jason Anderson
> > <Jason.Anderson@xxxxxxxxxxxxxxxx> wrote:
> >
> > Hello fellow Cephers,
> >
> >
> >
> > I am running Ceph Giant on Centos 6.6 and am running into an issue
> > where the create_pool() method in the librados python library isn't
> > using ceph.conf's pg_num and pgp_num settings during pool creation.
> >
> >
> >
> > My ceph.conf:
> >
> > [global]
> >
> > auth_service_required = cephx
> >
> > filestore_xattr_use_omap = true
> >
> > auth_client_required = cephx
> >
> > auth_cluster_required = cephx
> >
> > mon_host = 64.0.1.210,64.0.1.211
> >
> > public_network = 64.0.1.192/26
> >
> > mon_initial_members = MON05, MON06
> >
> > fsid = 91eda869-9a34-4f74-8189-7d5fb6952f4a
> >
> > osd pool default pg num = 256
> >
> > osd pool default pgp num = 256
> >
> >
> >
> > In python, using the above ceph.conf, conf_get returns the correct
> > pg_num and I am able to create the pool.
> >
> > [root@DCOS01 lib]# python
> >
> >>>> import rados
> >
> >>>> cluster = rados.Rados(conffile='/etc/ceph/ceph.conf')
> >
> >>>> cluster.conf_get('osd pool default pg num')
> >
> > '256'
> >
> >>>> cluster.connect()
> >
> >>>> cluster.create_pool('PE-TEST4')
> >
> >
> >
> > After the pool is created via the librados python library, I check the
> > pg_num of the pool and it's 8; not 256.
> >
> > [root@DCOS01 lib]# ceph osd pool get PE-TEST4 pg_num
> >
> > pg_num: 8
> >
> >
> >
> > Has anyone else run into this issue? Am I missing something? I know I
> > could just spawn a subprocess call to the ceph command line utility,
> > but I would like to avoid that in the name of a cleaner python
> > integration.
> >
> >
> >
> > Your assistance is greatly appreciated.
> >
> >
> >
> > Thank you,
> >
> > - Jason
> >
> >
> >
> > _______________________________________________
> > ceph-users mailing list
> > ceph-users@xxxxxxxxxxxxxx
> > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
> _______________________________________________
> ceph-users mailing list
> ceph-users@xxxxxxxxxxxxxx
> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com


-- 
Christian Balzer        Network/Systems Engineer                
chibi@xxxxxxx   	Global OnLine Japan/Fusion Communications
http://www.gol.com/
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com





[Index of Archives]     [Information on CEPH]     [Linux Filesystem Development]     [Ceph Development]     [Ceph Large]     [Ceph Dev]     [Linux USB Development]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [xfs]


  Powered by Linux