kuberntes: Disable the scale_manager for scale down

Due to bug #1746510, the kubernetes scale manager needs to be
disabled to not break the scale down command completely. Note, that when
magnum scales down the cluster will pick the nodes to scale randomly.

Related-Bug: #1746510
Change-Id: I8c3505ec6d155323288217e0c8ea54adabdff1c3
This commit is contained in:
Spyros Trigazis 2018-02-22 14:47:24 +00:00 committed by Spyros Trigazis (strigazi)
parent e2e19aaa97
commit de39220cac
4 changed files with 16 additions and 6 deletions

View File

@ -13,7 +13,6 @@
# under the License.
from magnum.drivers.common import k8s_monitor
from magnum.drivers.common.k8s_scale_manager import K8sScaleManager
from magnum.drivers.heat import driver
from magnum.drivers.k8s_coreos_v1 import template_def
@ -35,4 +34,7 @@ class Driver(driver.HeatDriver):
return k8s_monitor.K8sMonitor(context, cluster)
def get_scale_manager(self, context, osclient, cluster):
return K8sScaleManager(context, osclient, cluster)
# FIXME: Until the kubernetes client is fixed, remove
# the scale_manager.
# https://bugs.launchpad.net/magnum/+bug/1746510
return None

View File

@ -13,7 +13,6 @@
# under the License.
from magnum.drivers.common import k8s_monitor
from magnum.drivers.common.k8s_scale_manager import K8sScaleManager
from magnum.drivers.heat import driver
from magnum.drivers.k8s_fedora_atomic_v1 import template_def
@ -35,4 +34,7 @@ class Driver(driver.HeatDriver):
return k8s_monitor.K8sMonitor(context, cluster)
def get_scale_manager(self, context, osclient, cluster):
return K8sScaleManager(context, osclient, cluster)
# FIXME: Until the kubernetes client is fixed, remove
# the scale_manager.
# https://bugs.launchpad.net/magnum/+bug/1746510
return None

View File

@ -13,7 +13,6 @@
# under the License.
from magnum.drivers.common import k8s_monitor
from magnum.drivers.common.k8s_scale_manager import K8sScaleManager
from magnum.drivers.heat import driver
from magnum.drivers.k8s_fedora_ironic_v1 import template_def
@ -35,4 +34,7 @@ class Driver(driver.HeatDriver):
return k8s_monitor.K8sMonitor(context, cluster)
def get_scale_manager(self, context, osclient, cluster):
return K8sScaleManager(context, osclient, cluster)
# FIXME: Until the kubernetes client is fixed, remove
# the scale_manager.
# https://bugs.launchpad.net/magnum/+bug/1746510
return None

View File

@ -10,6 +10,10 @@ issues:
task uses the kubernetes client for kubernetes clusters and it crashes
the sync_cluster_status and send_cluster_metrics tasks.
https://bugs.launchpad.net/magnum/+bug/1746510
Additionally, the kubernetes scale manager needs to be disabled
to not break the scale down command completely. Note, that when magnum
scales down the cluster will pick the nodes to scale randomly.
upgrade:
- |
In magnum configuration, in [drivers] set send_cluster_metrics = False to