Define Resource Placement
The Workload Manager has ability to deploy enterprise applications over public, private, or hybrid clouds by configuring user-specified cloud settings in the Workload Manager UI > Environments > Edit Deployment Environment > Define Default Cloud Settings page.
The Resource Placement integration features extends the Workload Manager capabilities by allowing users to define cloud settings based on third-party infrastructure tools or quota management tools using automated scripts instead of manually-selected settings.
You can configure these integrations using an automation callout script.
Resource Placement Flow
To use the resource placement script, specify a URL to the script that you want to run. The resource placement script runs inside a Docker container in the CloudCenter Suite cluster (or Cloud Remote, if installed). You cannot pass in any custom values. Your script must output the values specified in this section.
The Resource Placement feature is only supported for AWS, VMware, and OpenStack clouds.
This script is executed for each Node launch (called for each VM). For example, if you have a single-tier application with the minimum number of nodes set to 2, then this script is executed twice – 1 tier x 2 nodes = 2 executions. However, the Workload Manager passes variables such as service types which allows you to decide where to place it in the VM.
Toggle the switch to YES in the Resource Placement section, as shown in the following screenshot.
If this feature is enabled, the Cloud Settings form in the Deployment Environments > Cloud Defaults page will be disabled.
Identify the script location and the specific script for the Resource Placement Configuration.
Be aware of the customizable options for AWS, OpenStack, and VMware:
- AWS Options
The following table describes AWS-specific cloud settings for the resource placement callout script.
AWS Setting Description vpcId The VPC for the node to be deployed. subnetId The subnet where the node should be deployed in the above VPC. securityGroupList The security groups where the node should be associated in the above VPC. vmTagsList The AWS tags to associate with the node. assignPublicIp Identifies if the node should be assigned with a public IP. nodeInfo Customizable node Information detail that is displayed in the Workload Manager UI Job Details Page for each node. If not provided, the Workload Manager generates the default nodeInfo based on the provided values.
- OpenStack Options
The following table describes OpenStack-specific cloud settings for the resource placement callout script.
OpenStack Setting Description TenantName The name of OpenStack tenant. zone The availability zone as described in Availability Sets and Zones > OpenStack. nicInfo The OpenStack network interface information. privateIPAllocationMode The private IP allocation strategy (DHCP or PREALLOCATE_IP). networkID The OpenStack Network ID. publicIpAllocate A flag to allocate the public IP address (Boolean: true/false). nodeInfo Customizable node Information detail that is displayed in the Workload Manager UI Job Details Page for each node. If not provided, the Workload Manager generates the default nodeInfo.
- VMware Options
The following table describes VMware-specific cloud settings for the resource placement callout script:
VMware Setting Description vmTagsList The AWS tags to associate with the node. UserDataCenterName The datacenter to deploy the node. UserClusterName The cluster to deploy the node in the above datacenter. UserResourcePoolName The resource pool used to deploy the node. UserDatastoreCluster The datastore cluster to associate with the node. UserFolderName The user folder used for the node deployment. RootDiskResizable
Identifies if the root disk is resizable (Boolean: true/false).
The Root Disk setting is available as a separate field and you can only select the Root Disk size if the Cloning Mode is set to Full Clone (Enable Full Clone = Selected). See VMware Network Settings for additional context.
FullClone Identifies if the node to be launched is with full clone (Boolean: true/false). SystemFolderName The folder from which the template is selected. networkList
The list of networks to attach to the node. The following table describes the format that various network types require. See VMware Configurations > Resource Placement for additional context.
VMware Network Format Required DVS and ACI networks
UserHost The ESX host to which the node is launched. nodeInfo
The following table describes available environment variables for the Resource Placement script.
Environment Variable Description Cloud eNV_cliqrAppTierName The tier name. All clouds, except Container Clouds and Cisco UCSD. CliqrTier_<tierName>_instanceType The Instance Type of the tier. eNV_imageName The image Name (for example: CentOS 6.x). serviceName The service name to identify settings like private subnet for a database service. eNV_parentJobName
The unique Job Name for the deployment.
CliqrCloudAccountId The cloud account ID. CliqrCloudAccountPwd The cloud account password (for AWS, access key). CliqrCloudAccountName The cloud account username (for AWS, account email ID). Cloud_Setting_CloudFamily The cloud family of the region in the Workload Manager. CliqrCloud_AccessSecretKey The AWS account secret key AWS
The SDK URL for VMware. VMware
The default OpenStack domain ID. OpenStack
The OpenStack Keystone authentication endpoint.
The OpenStack default tenant name.
The OpenStack default domain name.
The OpenStack password.
The OpenStack region.
The OpenStack user name for this account.
The OpenStack default tenant ID.
Container-Specific Resource Placement
A placement group represents a container pod and a tier in the application profile. A placement group is represented by a rectangle that you add to the topology modeler canvas by clicking the Create A Group button as indicated in the following screenshot. This support is available for some services like Apache, Nginx, MongoDB, and MySQL.
You can only drag and drop supported containers into the Create a Group rectangle.
Click and drag container-based services from the services palate on the left side of the topology modeler tab to the placement group rectangle, similar to the Apache2 profile and MongoDB profile that are visible in the canvas displayed in the following screenshot.
The properties panel on the right displays the properties for the currently selected object, which in this case is the last container dragged into the placement group, MongoDB. Notice that the minimum and maximum replicas fields in the General Settings section are not available, and the Network Services and Firewall Rules sections are also not available for this container. This is because these parameters are properties of the tier/pod. To see these parameters, click on the border of the placement group to select the placement group as a whole.
After selecting the placement group, notice the following in the Properties panel:
The name of the placement group contains the name of the first container dragged to the group (but you can edit this field).
Expanding the sections for Network Services and then Firewall Rules reveals that these sections automatically contain the union of the corresponding network services and firewall rules of the constituent containers.
The sections for Volumes, Deployment Parameters, and Minimum Resource Specifications are gone as these are specific to each container in the pod.
When deploying an application with a container tier that contains multiple containers, each container is listed within the corresponding per tier section on Page 1 of the deploy form.
If the container has visible parameters (defined in the application profile), an expand triangular icon appears to the left of the container name. Click on the expand icon to expand that section and display the deployment parameters.
On Page 2 of the deploy form you have the option to specify the instance type for each container separately.
After the application successfully deploys, the individual containers within each replica of the pod are shown in the deployment Details page as described in the Deployment Details section.
The following screenshot is an example of a Kubernetes console.
In the above screenshot:
The Port Name in the in the Workload Manager variable syntax (displayed in the following screenshot of the Workload Manager UI Properties pane) is the <service_name> in the Kubernetes console.
The string displayed in the Kubernetes console as http-668-335a5a is the longer unique name created by Workload Manager and passed to Kubernetes when the app is launched
See Pre-Defined Parameters > Kubernetes Container Service Parameters for additional context.
See Container Service for service-specific details.
- No labels