]>

Compute uses the nova-scheduler service to determine how to dispatch compute and volume requests. For example, the nova-scheduler service determines which host a VM should launch on. The term host in the context of filters means a physical node that has a nova-compute service running on it. You can configure the scheduler through a variety of options. Compute is configured with the following default scheduler options in the /etc/nova/nova.conf file: scheduler_driver=nova.scheduler.multi.MultiScheduler scheduler_driver_task_period=60 compute_scheduler_driver=nova.scheduler.filter_scheduler.FilterScheduler scheduler_available_filters=nova.scheduler.filters.all_filters scheduler_default_filters=RetryFilter,AvailabilityZoneFilter,RamFilter,ComputeFilter,ComputeCapabilitiesFilter,ImagePropertiesFilter,ServerGroupAntiAffinityFilter,ServerGroupAffinityFilter By default, the scheduler_driver is configured as a filter scheduler, as described in the next section. In the default configuration, this scheduler considers hosts that meet all the following criteria: Have not been attempted for scheduling purposes (RetryFilter). Are in the requested availability zone (AvailabilityZoneFilter). Have sufficient RAM available (RamFilter). Are capable of servicing the request (ComputeFilter). Satisfy the extra specs associated with the instance type (ComputeCapabilitiesFilter). Satisfy any architecture, hypervisor type, or virtual machine mode properties specified on the instance's image properties. (ImagePropertiesFilter). The scheduler caches its list of available hosts; you can specify how often the list is updated by modifying the scheduler_driver_task_period value. Do not configure service_down_time to be much smaller than scheduler_driver_task_period; otherwise, hosts will appear to be dead while the host list is being cached. For information on the volume scheduler, refer the Block Storage section of OpenStack Cloud Administrator Guide for information. The choice of a new host on instance migration is done by the scheduler. When evacuating instances from a host, the scheduler service does not pick the next host. Instances are evacuated to the host explicitly defined by the administrator. For information about instance evacuation, refer to the Evacuate instances section of the Cloud Administrator Guide.

The Filter Scheduler (nova.scheduler.filter_scheduler.FilterScheduler) is the default scheduler for scheduling virtual machine instances. It supports filtering and weighting to make informed decisions on where a new instance should be created.

When the Filter Scheduler receives a request for a resource, it first applies filters to determine which hosts are eligible for consideration when dispatching a resource. Filters are binary: either a host is accepted by the filter, or it is rejected. Hosts that are accepted by the filter are then processed by a different algorithm to decide which hosts to use for that request, described in the Weights section.

The scheduler_available_filters configuration option in nova.conf provides the Compute service with the list of the filters that are used by the scheduler. The default setting specifies all of the filter that are included with the Compute service: scheduler_available_filters = nova.scheduler.filters.all_filters This configuration option can be specified multiple times. For example, if you implemented your own custom filter in Python called myfilter.MyFilter and you wanted to use both the built-in filters and your custom filter, your nova.conf file would contain: scheduler_available_filters=nova.scheduler.filters.all_filters scheduler_available_filters=myfilter.MyFilter The scheduler_default_filters configuration option in nova.conf defines the list of filters that are applied by the nova-scheduler service. The default filters are: scheduler_default_filters=RetryFilter,AvailabilityZoneFilter,RamFilter,ComputeFilter,ComputeCapabilitiesFilter,ImagePropertiesFilter,ServerGroupAntiAffinityFilter,ServerGroupAffinityFilter The following sections describe the available filters.

Implements blueprint per-aggregate-resource-ratio. AggregateCoreFilter supports per-aggregate cpu_allocation_ratio. If the per-aggregate value is not found, the value falls back to the global setting.

Matches properties defined in an image's metadata against those of aggregates to determine host matches: If a host belongs to an aggregate and the aggregate defines one or more metadata that match an image's properties, that host is a candidate to boot the image's instance. If a host does not belong to any aggregate, it can boot instances from all images. For example, the following aggregate myWinAgg has the Windows operating system as metadata (named 'windows'): $ nova aggregate-details MyWinAgg +----+----------+-------------------+------------+---------------+ | Id | Name | Availability Zone | Hosts | Metadata | +----+----------+-------------------+------------+---------------+ | 1 | MyWinAgg | None | 'sf-devel' | 'os=windows' | +----+----------+-------------------+------------+---------------+ In this example, because the following Win-2012 image has the windows property, it would boot on the sf-devel host (all other filters being equal): $ glance image-show Win-2012 +------------------+--------------------------------------+ | Property | Value | +------------------+--------------------------------------+ | Property 'os' | windows | | checksum | f8a2eeee2dc65b3d9b6e63678955bd83 | | container_format | ami | | created_at | 2013-11-14T13:24:25 | | ... You can configure the AggregateImagePropertiesIsolation filter using the following options in the nova.conf file: # Considers only keys matching the given namespace (string). aggregate_image_properties_isolation_namespace=<None> # Separator used between the namespace and keys (string). aggregate_image_properties_isolation_separator=.

Matches properties defined in an instance type's extra specs against admin-defined properties on a host aggregate. Works with specifications that are unscoped, or are scoped with aggregate_instance_extra_specs. See the host aggregates section for documentation on how to use this filter.

Isolates tenants to specific host aggregates. If a host is in an aggregate that has the metadata key filter_tenant_id it only creates instances from that tenant (or list of tenants). A host can be in different aggregates. If a host does not belong to an aggregate with the metadata key, it can create instances from all tenants.

Implements blueprint per-aggregate-resource-ratio. Supports per-aggregate ram_allocation_ratio. If per-aggregate value is not found, it falls back to the default setting.

This is a no-op filter, it does not eliminate any of the available hosts.

Filters hosts by availability zone. This filter must be enabled for the scheduler to respect availability zones in requests.

Matches properties defined in an instance type's extra specs against compute capabilities. If an extra specs key contains a colon ":", anything before the colon is treated as a namespace, and anything after the colon is treated as the key to be matched. If a namespace is present and is not 'capabilities', it is ignored by this filter.

Passes all hosts that are operational and enabled. In general, this filter should always be enabled.

Only schedule instances on hosts if there are sufficient CPU cores available. If this filter is not set, the scheduler may over provision a host based on cores (for example, the virtual cores running on an instance may exceed the physical cores). This filter can be configured to allow a fixed amount of vCPU overcommitment by using the cpu_allocation_ratio Configuration option in nova.conf. The default setting is: cpu_allocation_ratio=16.0 With this setting, if 8 vCPUs are on a node, the scheduler allows instances up to 128 vCPU to be run on that node. To disallow vCPU overcommitment set: cpu_allocation_ratio=1.0 The Compute API will always return the actual number of CPU cores available on a compute node regardless of the value of the cpu_allocation_ratio configuration key. As a result changes to the cpu_allocation_ratio are not reflected via the command line clients or the dashboard. Changes to this configuration key are only taken into account internally in the scheduler.

Schedule the instance on a different host from a set of instances. To take advantage of this filter, the requester must pass a scheduler hint, using different_host as the key and a list of instance uuids as the value. This filter is the opposite of the SameHostFilter. Using the nova command-line tool, use the --hint flag. For example: $ nova boot --image cedef40a-ed67-4d10-800e-17455edce175 --flavor 1 \ --hint different_host=a0cf03a5-d921-4877-bb5c-86d26cf818e1 \ --hint different_host=8c19174f-4220-44f0-824a-cd1eeef10287 server-1 With the API, use the os:scheduler_hints key. For example:

Only schedule instances on hosts if there is sufficient disk space available for root and ephemeral storage. This filter can be configured to allow a fixed amount of disk overcommitment by using the disk_allocation_ratio Configuration option in nova.conf. The default setting is: disk_allocation_ratio=1.0 Adjusting this value to greater than 1.0 enables scheduling instances while over committing disk resources on the node. This might be desirable if you use an image format that is sparse or copy on write such that each virtual instance does not require a 1:1 allocation of virtual disk to physical storage.

This filter is deprecated in favor of ServerGroupAffinityFilter. The GroupAffinityFilter ensures that an instance is scheduled on to a host from a set of group hosts. To take advantage of this filter, the requester must pass a scheduler hint, using group as the key and an arbitrary name as the value. Using the nova command-line tool, use the --hint flag. For example: $ nova boot --image IMAGE_ID --flavor 1 --hint group=foo server-1 This filter should not be enabled at the same time as GroupAntiAffinityFilter or neither filter will work properly.

This filter is deprecated in favor of ServerGroupAntiAffinityFilter. The GroupAntiAffinityFilter ensures that each instance in a group is on a different host. To take advantage of this filter, the requester must pass a scheduler hint, using group as the key and an arbitrary name as the value. Using the nova command-line tool, use the --hint flag. For example: $ nova boot --image IMAGE_ID --flavor 1 --hint group=foo server-1 This filter should not be enabled at the same time as GroupAffinityFilter or neither filter will work properly.

Filters hosts based on properties defined on the instance's image. It passes hosts that can support the specified image properties contained in the instance. Properties include the architecture, hypervisor type, and virtual machine mode. for example, an instance might require a host that runs an ARM-based processor and QEMU as the hypervisor. An image can be decorated with these properties by using: $ glance image-update img-uuid --property architecture=arm --property hypervisor_type=qemu The image properties that the filter checks for are: architecture: Architecture describes the machine architecture required by the image. Examples are i686, x86_64, arm, and ppc64. hypervisor_type: Hypervisor type describes the hypervisor required by the image. Examples are xen, kvm, qemu, and xenapi. vm_mode: Virtual machine mode describes the hypervisor application binary interface (ABI) required by the image. Examples are 'xen' for Xen 3.0 paravirtual ABI, 'hvm' for native ABI, 'uml' for User Mode Linux paravirtual ABI, exe for container virt executable ABI.

Allows the admin to define a special (isolated) set of images and a special (isolated) set of hosts, such that the isolated images can only run on the isolated hosts, and the isolated hosts can only run isolated images. The flag restrict_isolated_hosts_to_isolated_images can be used to force isolated hosts to only run isolated images. The admin must specify the isolated set of images and hosts in the nova.conf file using the isolated_hosts and isolated_images configuration options. For example: isolated_hosts=server1,server2 isolated_images=342b492c-128f-4a42-8d3a-c5088cf27d13,ebd267a6-ca86-4d6c-9a0e-bd132d6b7d09

The JsonFilter allows a user to construct a custom filter by passing a scheduler hint in JSON format. The following operators are supported: = < > in <= >= not or and The filter supports the following variables: $free_ram_mb $free_disk_mb $total_usable_ram_mb $vcpus_total $vcpus_used Using the nova command-line tool, use the --hint flag: $ nova boot --image 827d564a-e636-4fc4-a376-d36f7ebe1747 \ --flavor 1 --hint query='[">=","$free_ram_mb",1024]' server1 With the API, use the os:scheduler_hints key:

Only schedule instances on hosts that have sufficient RAM available. If this filter is not set, the scheduler may over provision a host based on RAM (for example, the RAM allocated by virtual machine instances may exceed the physical RAM). This filter can be configured to allow a fixed amount of RAM overcommitment by using the ram_allocation_ratio configuration option in nova.conf. The default setting is: ram_allocation_ratio=1.5 This setting enables 1.5 GB instances to run on any compute node with 1 GB of free RAM.

Filter out hosts that have already been attempted for scheduling purposes. If the scheduler selects a host to respond to a service request, and the host fails to respond to the request, this filter prevents the scheduler from retrying that host for the service request. This filter is only useful if the scheduler_max_attempts configuration option is set to a value greater than zero.

Schedule the instance on the same host as another instance in a set of instances. To take advantage of this filter, the requester must pass a scheduler hint, using same_host as the key and a list of instance uuids as the value. This filter is the opposite of the DifferentHostFilter. Using the nova command-line tool, use the --hint flag: $ nova boot --image cedef40a-ed67-4d10-800e-17455edce175 --flavor 1 \ --hint same_host=a0cf03a5-d921-4877-bb5c-86d26cf818e1 \ --hint same_host=8c19174f-4220-44f0-824a-cd1eeef10287 server-1 With the API, use the os:scheduler_hints key:

The ServerGroupAffinityFilter ensures that an instance is scheduled on to a host from a set of group hosts. To take advantage of this filter, the requester must create a server group with an affinity policy, and pass a scheduler hint, using group as the key and the server group UUID as the value. Using the nova command-line tool, use the --hint flag. For example: $ nova server-group-create --policy affinity group-1 $ nova boot --image IMAGE_ID --flavor 1 --hint group=SERVER_GROUP_UUID server-1

The ServerGroupAntiAffinityFilter ensures that each instance in a group is on a different host. To take advantage of this filter, the requester must create a server group with an anti-affinity policy, and pass a scheduler hint, using group as the key and the server group UUID as the value. Using the nova command-line tool, use the --hint flag. For example: $ nova server-group-create --policy anti-affinity group-1 $ nova boot --image IMAGE_ID --flavor 1 --hint group=SERVER_GROUP_UUID server-1

Schedule the instance based on host IP subnet range. To take advantage of this filter, the requester must specify a range of valid IP address in CIDR format, by passing two scheduler hints: build_near_host_ip The first IP address in the subnet (for example, 192.168.1.1) cidr The CIDR that corresponds to the subnet (for example, /24) Using the nova command-line tool, use the --hint flag. For example, to specify the IP subnet 192.168.1.1/24 $ nova boot --image cedef40a-ed67-4d10-800e-17455edce175 --flavor 1 \ --hint build_near_host_ip=192.168.1.1 --hint cidr=/24 server-1 With the API, use the os:scheduler_hints key:

When resourcing instances, the Filter Scheduler filters and weighs each host in the list of acceptable hosts. Each time the scheduler selects a host, it virtually consumes resources on it, and subsequent selections are adjusted accordingly. This process is useful when the customer asks for the same large amount of instances, because weight is computed for each requested instance. All weights are normalized before being summed up; the host with the largest weight is given the highest priority.

If cells are used, cells are weighted by the scheduler in the same manner as hosts. Hosts and cells are weighed based on the following options in the /etc/nova/nova.conf file:

Host Weighting options

Section	Option	Description
[DEFAULT]	ram_weight_multiplier	By default, the scheduler spreads instances across all hosts evenly. Set the ram_weight_multiplier option to a negative number if you prefer stacking instead of spreading. Use a floating-point value.
[DEFAULT]	scheduler_host_subset_size	New instances are scheduled on a host that is chosen randomly from a subset of the N best hosts. This property defines the subset size from which a host is chosen. A value of 1 chooses the first host returned by the weighing functions.This value must be at least 1. A value less than 1 is ignored, and 1 is used instead. Use an integer value.
[DEFAULT]	scheduler_weight_classes	Defaults to nova.scheduler.weights.all_weighers, which selects the only available weigher, the RamWeigher. Hosts are then weighed and sorted with the largest weight winning.
[metrics]	weight_multiplier	Multiplier for weighing metrices. Use a floating-point value.
[metrics]	weight_setting	Determines how metrics are weighed. Use a comma-separated list of metricName=ratio. For example: "name1=1.0, name2=-1.0" results in: name1.value * 1.0 + name2.value * -1.0
[metrics]	required	Specifies how to treat unavailable metrics: True—Raises an exception. To avoid the raised exception, you should use the scheduler filter MetricFilter to filter out hosts with unavailable metrics. False—Treated as a negative factor in the weighing process (uses the weight_of_unavailable option).
[metrics]	weight_of_unavailable	If required is set to False, and any one of the metrics set by weight_setting is unavailable, the weight_of_unavailable value is returned to the scheduler.

For example: [DEFAULT] scheduler_host_subset_size=1 scheduler_weight_classes=nova.scheduler.weights.all_weighers ram_weight_multiplier=1.0 [metrics] weight_multiplier=1.0 weight_setting=name1=1.0, name2=-1.0 required=false weight_of_unavailable=-10000.0

Cell weighting options

Section	Option	Description
[cells]	mute_weight_multiplier	Multiplier to weigh mute children (hosts which have not sent capacity or capacity updates for some time). Use a negative, floating-point value.
[cells]	mute_weight_value	Weight value assigned to mute children. Use a positive, floating-point value with a maximum of '1.0'.
[cells]	offset_weight_multiplier	Multiplier to weigh cells, so you can specify a preferred cell. Use a floating point value.
[cells]	ram_weight_multiplier	By default, the scheduler spreads instances across all cells evenly. Set the ram_weight_multiplier option to a negative number if you prefer stacking instead of spreading. Use a floating-point value.
[cells]	scheduler_weight_classes	Defaults to nova.cells.weights.all_weighers, which maps to all cell weighers included with Compute. Cells are then weighed and sorted with the largest weight winning.

For example: [cells] scheduler_weight_classes=nova.cells.weights.all_weighers mute_weight_multiplier=-10.0 mute_weight_value=1000.0 ram_weight_multiplier=1.0 offset_weight_multiplier=1.0

As an administrator, you work with the Filter Scheduler. However, the Compute service also uses the Chance Scheduler, nova.scheduler.chance.ChanceScheduler, which randomly selects from lists of filtered hosts.

To customize the Compute scheduler, use the configuration option settings documented in .