openstack/openstack-helm-infra
Code Issues Proposed changes

[ceph-pools] Enable autoscaler by reading from values

Browse Source 
This is to make enabling autosclaer feature optional from values since
its new feature and few deployments may not required it.

Change-Id: Ie7cbdb71f6a1bf636db2d61ae0f3a4c19af2ca34
This commit is contained in:
Chinasubbareddy Mallavarapu 2020-03-28 15:51:20 -05:00
parent 13f54b0e03
commit 872f68ada7
3 changed files with 9 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
ceph-client/templates/bin/pool/_init.sh.tpl
View File

@ -77,8 +77,10 @@ function reweight_osds () {
}
function enable_autoscaling () {
ceph mgr module enable pg_autoscaler
ceph config set global osd_pool_default_pg_autoscale_mode on
if [[ "${ENABLE_AUTOSCALER}" == "true" ]]; then
ceph mgr module enable pg_autoscaler
ceph config set global osd_pool_default_pg_autoscale_mode on
fi
}
function create_pool () {
@ -93,7 +95,7 @@ function create_pool () {
while [ $(ceph --cluster "${CLUSTER}" -s | grep creating -c) -gt 0 ]; do echo -n .;sleep 1; done
ceph --cluster "${CLUSTER}" osd pool application enable "${POOL_NAME}" "${POOL_APPLICATION}"
else
if [[ -z "$(ceph osd versions | grep ceph\ version | grep -v nautilus)" ]]; then
if [[ -z "$(ceph osd versions | grep ceph\ version | grep -v nautilus)" ]] && [[ $"{ENABLE_AUTOSCALER}" == "true" ]] ; then
ceph --cluster "${CLUSTER}" osd pool set "${POOL_NAME}" pg_autoscale_mode on
fi
fi
@ -155,13 +157,7 @@ function manage_pool () {
POOL_PROTECTION=$8
CLUSTER_CAPACITY=$9
TOTAL_OSDS={{.Values.conf.pool.target.osd}}
# This is a workaround for a pg merging bug in ceph. The only reason this works is because the
# autoscaler is set to work unconditionally. This needs to change once the autoscaler is optional.
if [[ -z "$(ceph osd versions | grep ceph\ version | grep -v nautilus)" ]]; then
POOL_PLACEMENT_GROUPS=4
else
POOL_PLACEMENT_GROUPS=$(/tmp/pool-calc.py ${POOL_REPLICATION} ${TOTAL_OSDS} ${TOTAL_DATA_PERCENT} ${TARGET_PG_PER_OSD})
fi
POOL_PLACEMENT_GROUPS=$(/tmp/pool-calc.py ${POOL_REPLICATION} ${TOTAL_OSDS} ${TOTAL_DATA_PERCENT} ${TARGET_PG_PER_OSD})
create_pool "${POOL_APPLICATION}" "${POOL_NAME}" "${POOL_REPLICATION}" "${POOL_PLACEMENT_GROUPS}" "${POOL_CRUSH_RULE}" "${POOL_PROTECTION}"
POOL_REPLICAS=$(ceph --cluster "${CLUSTER}" osd pool get "${POOL_NAME}" size | awk '{print $2}')
POOL_QUOTA=$(python -c "print(int($CLUSTER_CAPACITY * $TOTAL_DATA_PERCENT * $TARGET_QUOTA / $POOL_REPLICAS / 100 / 100))")

2
ceph-client/templates/job-rbd-pool.yaml
View File

@ -52,6 +52,8 @@ spec:
env:
- name: CLUSTER
value: "ceph"
- name: ENABLE_AUTOSCALER
value: {{ .Values.conf.features.pg_autoscaler | quote }}
command:
- /tmp/pool-init.sh
volumeMounts:

1
ceph-client/values.yaml
View File

@ -230,6 +230,7 @@ conf:
features:
mds: true
mgr: true
pg_autoscaler: true
pool:
#NOTE(portdirect): this drives a simple approximation of
# https://ceph.com/pgcalc/, the `target.osd` key should be set to match the