Deploying Tanzu Application Services on Dell EMC PowerFlex
Tue, 15 Dec 2020 14:35:58 -0000
|Read Time: 0 minutes
Introduction
Tanzu Application Service (TAS) architecture provides the best approach available today to enable agility at scale with the reliability that is must to address these challenges. PowerFlex family offers key value propositions of traditional and cloud-native production workloads, deployment flexibility, linear scalability, predictable high performance, and enterprise-grade resilience.
Tanzu Application Service (TAS)
The VMware Tanzu Application Service (TAS) is based on Cloud Foundry –an open-source cloud application platform that provides a choice of clouds, developer frameworks, and application services. Cloud Foundry is a multi-cloud platform for the deployment, management, and continuous delivery of applications, containers, and functions. TAS abstracts away the process of setting up and managing an application runtime environment so that developers can focus solely on their applications and associated data. Running a single command—cf push—creates a scalable environment for your application in seconds, which might otherwise take hours to spin up manually. TAS allows developers to deploy and deliver software quickly, without the need of managing the underlying infrastructure.
PowerFlex
PowerFlex (previously VxFlex OS) is the software foundation of PowerFlex software-defined storage. It is a unified compute, storage and networking solution delivering scale-out block storage service designed to deliver flexibility, elasticity, and simplicity with predictable high performance and resiliency at scale.
The PowerFlex platform is available in multiple consumption options to help customers meet their project and data center requirements. PowerFlex appliance and PowerFlex rack provide customers comprehensive IT Operations Management (ITOM) and life cycle management (LCM) of the entire infrastructure stack in addition to sophisticated high-performance, scalable, resilient storage services. PowerFlex appliance and PowerFlex rack are the two preferred and proactively marketed consumption options. PowerFlex is also available on VxFlex Ready Nodes for those customers interested in software-defined compliant hardware without the ITOM and LCM capabilities.
PowerFlex software-define storage with unified compute and networking offers flexibility of deployment architecture to help best meet the specific deployment and architectural requirements. PowerFlex can be deployed in a two-layer for asymmetrical scaling of compute and storage for “right-sizing capacities, single-layer (HCI), or in mixed architecture.
Deploying TAS on PowerFlex
For this example, a PowerFlex production cluster is set up using a Hyperconverged configuration. The production cluster has connectivity to the customer-data network and the private backend PowerFlex storage network. The PowerFlex production cluster consists of a minimum of four servers that host the workload and PowerFlex storage VMs. All the nodes are part of a single ESXi Cluster and part of the same PowerFlex Cluster. Each node contributes all their internal disk resources to PowerFlex cluster.
The PowerFlex management software manages the capacity of all of the disks and acts as a back-end for data access by presenting storage volumes to be consumed by the applications running on the nodes. PowerFlex Manager also provides the essential operational controls and lifecycle management tools. The production cluster hosts the compute nodes that are used for deployment of TAS VMs. TAS components are deployed across three dedicated compute clusters that are designated as three availability zones. These compute clusters are managed by the same 'compute workload' vCenter as the dedicated Edge cluster. The following figure depicts the layout in the lab environment:
Figure 1. PowerFlex production cluster
The compute infrastructure illustrates the best practice architecture using 3 AZ’s using PowerFlex rack in hyperconverged configured nodes. This design ensures the high availability of nodes (i.e., nodes in AZ1 will still function if AZ2 or AZ3 goes down). A dedicated compute cluster in each AZ’s combines to form Isolation Zone (IZ). These AZ’s can be used to deploy and run the TAS stateful workloads requiring persistent storage. On the PowerFlex storage we have created volumes in the backend which are being mapped to vSphere as Datastores.
PowerFlex storage distributed data layout scheme is designed to maximize protection and optimize performance. A single volume is divided into chunks. These chunks will be distributed (striped) on physical disks throughout the cluster, in a balanced and random manner. Each chunk has a total of two copies for redundancy.
PowerFlex can be feature configured optionally to achieve additional data redundancy by enabling the feature Fault sets. Persistent Storage for each AZ could be its own PowerFlex cluster. By implementing PowerFlex feature Fault sets we can ensure that the persistent data availability all time. Fault Sets are subgroup of SDS s (Software defined Storage) installed on host servers within a Protection Domain. PowerFlex OS will mirror data for a Fault Set on SDSs that are outside the Fault Set. Thus, availability is assured even if all the servers within one Fault Set fail simultaneously.
PowerFlex enables flexible scale out capabilities for your data center also provides unparalleled elasticity and scalability. Start with a small environment for your proof of concept or a new application and add nodes as needed when requirements evolve.
The solution mentioned in this blog provides recommendations for deploying a highly available and production-ready Tanzu Application Service on Dell EMC PowerFlex rack infrastructure platform to meet the performance, scalability, resiliency, and availability requirements and describes its hardware and software components. For complete information, see Tanzu Application Services on PowerFlex rack - Solution Guide.
References
Related Blog Posts
Take VMware Tanzu to the Cloud Edge with Dell Technologies Cloud Platform
Mon, 02 Nov 2020 15:50:28 -0000
|Read Time: 0 minutes
Dell Technologies and VMware are happy to announce the availability of VMware Cloud Foundation 4.1.0 on VxRail 7.0.100.
This release brings support for the latest versions of VMware Cloud Foundation and Dell EMC VxRail to the Dell Technologies Cloud Platform and provides a simple and consistent operational experience for developer ready infrastructure across core, edge, and cloud. Let’s review these new features.
Updated VMware Cloud Foundation and VxRail BOM
Cloud Foundation 4.1 on VxRail 7.0.100 introduces support for the latest versions of the SDDC listed below:
- vSphere 7.0 U1
- vSAN 7.0 U1
- NSX-T 3.0 P02
- vRealize Suite Lifecycle Manager 8.1 P01
- vRealize Automation 8.1 P02
- vRealize Log Insight 8.1.1
- vRealize Operations Manager 8.1.1
- VxRail 7.0.100
For the complete list of component versions in the release, please refer to the VCF on VxRail release notes. A link is available at the end of this post.
VMware Cloud Foundation Software Feature Updates
VCF on VxRail Management Enhancements
vSphere Cluster Level Services (vCLS)
vSphere Cluster Services is a new capability introduced in the vSphere 7 Update 1 release that is included as a part of VCF 4.1. It runs as a set of virtual machines deployed on top of every vSphere cluster. Its initial functionality provides foundational capabilities that are needed to create a decoupled and distributed control plane for clustering services in vSphere. vCLS ensures cluster services like vSphere DRS and vSphere HA are all available to maintain the resources and health of the workloads running in the clusters independent of the availability of vCenter Server. The figure below shows the components that make up vCLS from the vSphere Web Client.
Figure 1
Not only is vSphere 7 providing modernized data services like embedded vSphere Native Pods with vSphere with Tanzu but features like vCLS are now beginning the evolution of modernizing to distributed control planes too!
VCF Managed Resources and VxRail Cluster Object Renaming Support
VCF can now rename resource objects post creation, including the ability to rename domains, datacenters, and VxRail clusters.
The domain is managed by the SDDC Manager. As a result, you will find that there are additional options within the SDDC Manager UI that will allow you to rename these objects.
VxRail Cluster objects are managed by a given vCenter server instance. In order to change cluster names, you will need to change the name within vCenter Server. Once you do, you can go back to the SDDC Manager and after a refresh of the UI, the new cluster name will be retrieved by the SDDC Manager and shown.
In addition to the domain and VxRail cluster object rename, SDDC Manager now supports the use of a customized Datacenter object name. The enhanced VxRail VI WLD creation wizard process has been updated to include inputs for Datacenter Name and is automatically imported into the SDDC Manager inventory during the VxRail VI WLD Creation SDDC Manager workflow. Note: Make sure the Datacenter name matches the one used during the VxRail Cluster First Run. The figure below shows the Datacenter Input step in the enhanced VxRail VI WLD creation wizard from within SDDC Manager.
Figure 2
Being able to customize resource object names makes VCF on VxRail more flexible in aligning with an IT organization’s naming policies.
VxRail Integrated SDDC Manager WLD Cluster Node Removal Workflow Optimization
Furthering the Dell Technologies and VMware co-engineering integration efforts for VCF on VxRail, new workflow optimizations have been introduced in VCF 4.1 that take advantage of VxRail Manager APIs for VxRail cluster host removal operations.
When the time comes for VCF on VxRail cloud administrators to remove hosts from WLD clusters and repurpose them for other domains, admins will use the SDDC Manager “Remove Host from WLD Cluster” workflow to perform this task. This remove host operation has now been fully integrated with native VxRail Manager APIs to automate removing physical VxRail hosts from a VxRail cluster as a single end-to-end automated workflow that is kicked off from the SDDC Manager UI or VCF API. This integration further simplifies and streamlines VxRail infrastructure management operations all from within common VMware SDDC management tools. The figure below illustrates the SDDC Manager sub tasks that include new VxRail API calls used by SDDC Manager as a part of the workflow.
Figure 3
Note: Removed VxRail nodes require reimaging prior to repurposing them into other domains. This reimaging currently requires Dell EMC support to perform.
I18N Internationalization and Localization (SDDC Manager)
SDDC Manager now has international language support that meets the I18N Internationalization and Localization standard. Options to select the desired language are available in the Cloud Builder UI, which installs SDDC Manager using the selected language settings. SDDC Manager will have localization support for the following languages – German, Japanese, Chinese, French, and Spanish. The figure below illustrates an example of what this would look like in the SDDC Manager UI.
Figure 4
vRealize Suite Enhancements
VCF Aware vRSLCM
New in VCF 4.1, the vRealize Suite is fully integrated into VCF. The SDDC Manager deploys the vRSLCM and creates a two way communication channel between the two components. When deployed, vRSLCM is now VCF aware and reports back to the SDDC Manager what vRealize products are installed. The installation of vRealize Suite components utilizes built standardized VVD best practices deployment designs leveraging Application Virtual Networks (AVNs).
Software Bundles for the vRealize Suite are all downloaded and managed through the SDDC Manager. When patches or updates become available for the vRealize Suite, lifecycle management of the vRealize Suite components is controlled from the SDDC Manager, calling on vRSLCM to execute the updates as part of SDDC Manager LCM workflows. The figure below showcases the process for enabling vRealize Suite for VCF.
Figure 5
VCF Multi-Site Architecture Enhancements
VCF Remote Cluster Support
VCF Remote Cluster Support enables customers to extend their VCF on VxRail operational capabilities to ROBO and Cloud Edge sites, enabling consistent operations from core to edge. Pair this with an awesome selection of VxRail hardware platform options and Dell Technologies has your Edge use cases covered. More on hardware platforms later…For a great detailed explanation on this exciting new feature check out the link to a detailed VMware blog post on the topic at the end of this post.
VCF LCM Enhancements
NSX-T Edge and Host Cluster-Level and Parallel Upgrades
With previous VCF on VxRail releases, NSX-T upgrades were all encompassing, meaning that a single update required updates to all the transport hosts as well as the NSX Edge and Manager components in one evolution.
With VCF 4.1, support has been added to perform staggered NSX updates to help minimize maintenance windows. Now, an NSX upgrade can consist of three distinct parts:
- Updating of edges
- Can be one job or multiple jobs. Rerun the wizard.
- Must be done before moving to the hosts
- Updating the transport hosts
- Once the hosts within the clusters have been updated, the NSX Managers can be updated.
Multiple NSX edge and/or host transport clusters within the NSX-T instance can be upgraded in parallel. The Administrator has the option to choose some clusters without having to choose all of them. Clusters within a NSX-T fabric can also be chosen to be upgraded sequentially, one at a time. Below are some examples of how NSX-T components can be updated.
NSX-T Components can be updated in several ways. These include updating:
- NSX-T Edges and Host Clusters within an NSX-T instance can be upgraded together in parallel (default)
- NSX-T Edges can be upgraded independently of NSX-T Host Clusters
- NSX-T Host Clusters can be upgraded independently of NSX-T Edges only after the Edges are upgraded first
- NSX-T Edges and Host Clusters within an NSX-T instance can be upgraded sequentially one after another.
The figure below visually depicts these options.
Figure 6
These options provide Cloud admins with a ton of flexibility so they can properly plan and execute NSX-T LCM updates within their respective maintenance windows. More flexible and simpler operations. Nice!
VCF Security Enhancements
Read-Only Access Role, Local and Service Accounts
A new ‘view-only’ role has been added to VCF 4.1. For some context, let’s talk a bit now about what happens when logging into the SDDC Manager.
First, you will provide a username and password. This information gets sent to the SDDC Manager, who then sends it to the SSO domain for verification. Once verified, the SDDC Manager can see what role the account has privilege for.
In previous versions of Cloud Foundation, the role would either be for an Administrator or it would be for an Operator.
Now, there is a third role available called a ‘Viewer’. Like its name suggests, this is a view only role which has no ability to create, delete, or modify objects. Users who are assigned this role may not see certain items in the SDDC Manger UI, such as the User screen. They may also see a message saying they are unauthorized to perform certain actions.
Also new, VCF now has a local account that can be used during an SSO failure. To help understand why this is needed let’s consider this: What happens when the SSO domain is unavailable for some reason? In this case, the user would not be able to login. To address this, administrators now can configure a VCF local account called admin@local. This account will allow the performing of certain actions until the SSO domain is functional again. This VCF local account is defined in the deployment worksheet and used in the VCF bring up process. If bring up has already been completed and the local account was not configured, then a warning banner will be displayed on the SDDC Manager UI until the local account is configured.
Lastly, SDDC Manager now uses new service accounts to streamline communications between SDDC manager and the products within Cloud Foundation. These new service accounts follow VVD guidelines for pre-defined usernames and are administered through the admin user account to improve inter-VCF communications within SDDC Manager.
VCF Data Protection Enhancements
As described in this blog, with VCF 4.1, SDDC Manager backup-recovery workflows and APIs have been improved to add capabilities such as backup management, backup scheduling, retention policy, on-demand backup & auto retries on failure. The improvement also includes Public APIs for 3rd party ecosystem and certified backup solutions from Dell PowerProtect.
VxRail Software Feature Updates
VxRail Networking Enhancements
VxRail 4 x 25Gbps pNIC redundancy
VxRail engineering continues innovate in areas that drive more value to customers. The latest VCF on VxRail release follows through on delivering just that for our customers. New in this release, customers can use the automated VxRail First Run Process to deploy VCF on VxRail nodes using 4 x 25Gbps physical port configurations to run the VxRail System vDS for system traffic like Management, vSAN, and vMotion, etc. The physical port configuration of the VxRail nodes would include 2 x 25Gbps NDC ports and additional 2 x 25Gbps PCIe NIC ports.
In this 4 x 25Gbps set up, NSX-T traffic would run on the same System vDS. But what is great here (and where the flexibility comes in) is that customers can also choose to separate NSX-T traffic on its own NSX-T vDS that uplinks to separate physical PCIe NIC ports by using SDDC Manager APIs. This ability was first introduced in the last release and can also be leveraged here to expand the flexibility of VxRail host network configurations.
The figure below illustrates the option to select the base 4 x 25Gbps port configuration during VxRail First Run.
Figure 7
By allowing customers to run the VxRail System VDS across the NDC NIC ports and PCIe NIC ports, customers gain an extra layer of physical NIC redundancy and high availability. This has already been supported with 10Gbps based VxRail nodes. This release now brings the same high availability option to 25Gbps based VxRail nodes. Extra network high availability AND 25Gbps performance!? Sign me up!
VxRail Hardware Platform Updates
Recently introduced support for ruggedized D-Series VxRail hardware platforms (D560/D560F) continue expanding the available VxRail hardware platforms supported in the Dell Technologies Cloud Platform.
These ruggedized and durable platforms are designed to meet the demand for more compute, performance, storage, and more importantly, operational simplicity that deliver the full power of VxRail for workloads at the edge, in challenging environments, or for space-constrained areas.
These D-Series systems are a perfect match when paired with the latest VCF Remote Cluster features introduced in Cloud Foundation 4.1.0 to enable Cloud Foundation with Tanzu on VxRail to reach these space-constrained and challenging ROBO/Edge sites to run cloud native and traditional workloads, extending existing VCF on VxRail operations to these locations! Cool right?!
To read more about the technical details of VxRail D-Series, check out the VxRail D-Series Spec Sheet.
Well that about covers it all for this release. The innovation train continues. Until next time, feel free to check out the links below to learn more about DTCP (VCF on VxRail).
Jason Marques
Twitter - @vwhippersnapper
Additional Resources
VMware Blog Post on VCF Remote Clusters
Cloud Foundation on VxRail Release Notes
VxRail page on DellTechnologies.com
VCF on VxRail Interactive Demos
Introducing the PowerFlex Management Pack for vRealize Operations
Mon, 02 Nov 2020 13:09:42 -0000
|Read Time: 0 minutes
By Vineeth A C
Achieving operation efficiency in today’s modern cloud infrastructure brings automation to the forefront. Centralized visibility provides a key piece of the insight needed to understand if there are operational inefficiencies for taking actions that mitigate business disruption.
We are pleased to share the general availability of Dell EMC PowerFlex Management Pack for vRealize Operations 8.x. The PowerFlex MP for vROps extends the visibility of PowerFlex systems into vROps where IT can monitor their complete data center and cloud operations. It is available to all PowerFlex rack and appliance customers at no additional cost. This brings additional value to the comprehensive IT operations management functionality delivered by PowerFlex Manager that enables full life cycle management of the unified compute and software defined storage solution.
The management pack queries and collects key PowerFlex metrics for storage, compute, networking, and server hardware using APIs and ingests into vROps that can be visualized using the out-of-the-box dashboards. It also provides a detailed system level view that shows the health status and relationship between different components of the PowerFlex system.
Key features and capabilities
Dashboards: The management pack includes 13 default dashboards showing details of PowerFlex storage, PowerFlex Manager, PowerFlex nodes, network switches, ESXi hosts, and clusters. These configurable dashboards provide user customizable data displays that adjust to meet a wide variety of requirements.
Predefined symptoms and alert definitions: The management pack includes 166 symptom definitions and 152 alert definitions based on engineering best practices for the PowerFlex systems. Symptoms and alerts can be customized by the user to meet the demand of their environment.
Historical data: This is available for all PowerFlex Adapter resource kinds. This data provides a view of consumption over time and includes capacity forecasting based on usage for PowerFlex storage.
Network topology and relationship: The topology tree functionality available in vROps is extremely useful when mapping relationships between nodes, network interfaces, switch port, VLAN, port-channel, and vPC.
Detailed metric collection: In addition to the default dashboards, users have the option of drilling into specific metrics for nearly all available data from the components of PowerFlex system, even if it is not included in a dashboard.
Multiple PowerFlex systems awareness: Ability to group and differentiate multiple PowerFlex systems.
PowerFlex node type differentiation: Ability to identify and classify compute, storage, hyperconverged, and management controller nodes.
Sample dashboards
PowerFlex Details: This dashboard shows all the PowerFlex storage KPIs with historical data providing a view of storage performance utilization over time.
PowerFlex Node Summary: You can monitor the health status of all your PowerFlex nodes and its hardware components in this dashboard.
PowerFlex Networking Performance: This dashboard shows network KPIs like throughput, errors, packet discards with historical data providing a view of network utilization over time.
For customers who have already invested in vRealize Operations, this management pack is a great value add to monitor their PowerFlex systems. It is an end-to-end monitoring and alerting solution for PowerFlex infrastructure using vROps. It helps customers significantly in terms of capacity planning based on the historical data of resource consumption over time. It also helps to identify usage trends and provides insight to understand if there are operational issues/ inefficiencies for taking necessary actions to avoid service outages and mitigate business disruption. This integration with VMware vRealize Operations reduces operational complexity by using a unified platform to monitor and manage private data center infrastructure, as well as hybrid and multi-cloud environments.
References
- Download the PowerFlex Management Pack from the Flexera portal.
- Visit Infohub for product documentation.
- Visit PowerFlex site for complete information about PowerFlex software-defined storage.