openstack/kolla-ansible
Code Issues Proposed changes

Reduce the default values for Ceph pgs

Browse Source 
Required to keep Ceph working once we move to Luminous 12.2.1

Change-Id: I8d3e56f2053c939ea313c60cc04c0ff79dd27d25
Closes-Bug: 1763356
This commit is contained in:
Paul Bourke 2018-04-24 14:07:25 +01:00
parent cc6960dd04
commit 36f33f089b
4 changed files with 35 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
ansible/group_vars/all.yml
View File

@ -723,8 +723,11 @@ ceph_rule: "default host {{ 'indep' if ceph_pool_type == 'erasure' else 'firstn'
ceph_cache_rule: "cache host firstn" ceph_cache_rule: "cache host firstn"
# Set the pgs and pgps for pool # Set the pgs and pgps for pool
ceph_pool_pg_num: 128 # WARNING! These values are dependant on the size and shape of your cluster -
ceph_pool_pgp_num: 128 # the default values are not suitable for production use. Please refer to the
# Kolla Ceph documentation for more information.
ceph_pool_pg_num: 8
ceph_pool_pgp_num: 8
##################### #####################
# VMware support # VMware support

12
doc/source/reference/ceph-guide.rst
View File

@ -126,6 +126,18 @@ RadosGW is optional, enable it in ``/etc/kolla/globals.yml``:
.. end .. end
.. note::
Regarding number of placement groups (PGs)
Kolla sets very conservative values for the number of PGs per pool
(`ceph_pool_pg_num` and `ceph_pool_pgp_num`). This is in order to ensure
the majority of users will be able to deploy Ceph out of the box. It is
*highly* recommended to consult the official Ceph documentation regarding
these values before running Ceph in any kind of production scenario.
.. end
RGW requires a healthy cluster in order to be successfully deployed. On initial RGW requires a healthy cluster in order to be successfully deployed. On initial
start up, RGW will create several pools. The first pool should be in an start up, RGW will create several pools. The first pool should be in an
operational state to proceed with the second one, and so on. So, in the case of operational state to proceed with the second one, and so on. So, in the case of

7
etc/kolla/globals.yml
View File

@ -267,8 +267,11 @@ kolla_internal_vip_address: "10.10.10.254"
#enable_ceph_rgw_keystone: "no" #enable_ceph_rgw_keystone: "no"
# Set the pgs and pgps for pool # Set the pgs and pgps for pool
#ceph_pool_pg_num: 128 # WARNING! These values are dependant on the size and shape of your cluster -
#ceph_pool_pgp_num: 128 # the default values are not suitable for production use. Please refer to the
# Kolla Ceph documentation for more information.
#ceph_pool_pg_num: 8
#ceph_pool_pgp_num: 8
############################# #############################
# Keystone - Identity Options # Keystone - Identity Options

13
releasenotes/notes/reduce-ceph-pgs-27e88e3b6e3b809c.yaml Normal file
View File

@ -0,0 +1,13 @@
---
issues:
- |
As of Ceph Luminous 12.2.1 the maximum number of PGs per OSD before the
monitor issues a warning has been reduced from 300 to 200 PGs. In addition,
Ceph now fails with an error rather than a warning in the case of exeeding
the max value.
In order to allow Kolla to continue to be used out of the box we have
reduced the default values for pg_num and pgp_num from 128 to 8. This will
allow a deploy of Kolla with all possible services enabled and then some,
with the minimum recommended three OSDs. Operators are *highly*
recommended to review the Ceph documentation regarding these values in
order to ensure optimal performance for their own cluster.