Bases: watcher.decision_engine.strategy.strategies.base.WorkloadStabilizationBaseStrategy
Workload Stabilization control using live migration
Description
This is workload stabilization strategy based on standard deviation
algorithm. The goal is to determine if there is an overload in a cluster
and respond to it by migrating VMs to stabilize the cluster.
Requirements
- Software: Ceilometer component ceilometer-compute running
in each compute host, and Ceilometer API can report such telemetries
memory.resident and cpu_util successfully.
- You must have at least 2 physical compute nodes to run this strategy.
Limitations
- It assume that live migrations are possible
- Load on the system is sufficiently stable.
Spec URL
https://review.openstack.org/#/c/286153/
-
MEMOIZE(fn)
-
add_migration(resource_id, migration_type, source_node, destination_node)[source]
-
calculate_migration_case(hosts, instance_id, src_node_id, dst_node_id)[source]
Calculate migration case
Return list of standard deviation values, that appearing in case of
migration of instance from source host to destination host
:param hosts: hosts with their workload
:param instance_id: the virtual machine
:param src_node_id: the source node id
:param dst_node_id: the destination node id
:return: list of standard deviation values
-
calculate_weighted_sd(sd_case)[source]
Calculate common standard deviation among meters on host
-
check_threshold()[source]
Check if cluster is needed in balancing
-
create_migration_instance(mig_instance, mig_source_node, mig_destination_node)[source]
Create migration VM
-
do_execute()[source]
-
fill_solution()[source]
-
classmethod get_display_name()[source]
-
get_hosts_load()[source]
Get load of every host by gathering instances load
-
get_instance_load(*arg, **kw)[source]
Gathering instance load through ceilometer statistic.
| Parameters: | instance_uuid – instance for which statistic is gathered. |
|---|
| Returns: | dict |
|---|
-
classmethod get_name()[source]
-
classmethod get_schema()[source]
-
get_sd(hosts, meter_name)[source]
Get standard deviation among hosts by specified meter
-
classmethod get_translatable_display_name()[source]
-
migrate(instance_uuid, src_host, dst_host)[source]
-
normalize_hosts_load(hosts)[source]
-
post_execute()[source]
Post-execution phase
This can be used to compute the global efficacy
-
pre_execute()[source]
-
simulate_migrations(hosts)[source]
Make sorted list of pairs instance:dst_host
-
transform_instance_cpu(instance_load, host_vcpus)[source]
Transform instance cpu utilization to overall host cpu utilization.
| Parameters: |
- instance_load – dict that contains instance uuid and
utilization info.
- host_vcpus – int
|
|---|
| Returns: | float value
|
|---|