In the previous posts about PoweFlex 3.5, we have covered the new HTML5 UI and the new replication features which are part of PowerFlex 3.5

In this post, we will provide an overview of the upper level element manager which is used to manage both the storage part AND the servers / networking part which is used when you purchase PowerFlex as an ‘Appliance’ or a ‘Rack’ (https://www.delltechnologies.com/he-il/storage/powerflex.htm)

PowerFlex Manager is meant to provide a robust toolset that helps our customers automate IT operations and lifecycle management tasks.

It was designed to help automate operational tasks across storage and compute, which reduces the steps involved in routine operations and helps simplify compliance activities.​

​It also provides templates that orchestrate activities across infrastructure layers, which increases overall IT productivity.

​And…PowerFlex Manager provides APIs to help integrate PowerFlex infrastructure with DevOps and cloud delivery workflows.​

PowerFlex Manager key functions

Built to apply ITOM principals to managing and operating the PowerFlex family

  • Unified management and automation for PowerFlex Family
  • Hardware and software monitoring and alerting
  • System deployment and operations
  • Template-based provisioning of infrastructure and workloads
  • Simplified, automated lifecycle management
  • Monitoring, alerting and health status of all components
  • Orchestrates software updates based on catalog validation

The PowerFlex Manager tool uses IT Operations Management (ITOM) Principles for its framework. ITOM is covered under the IT Infrastructure Library (ITIL) framework and sets the industry standard for IT operations management. It focuses on the execution of day-to-day tasks and lifecycle management. It is a great way to understand the needs of an IT organization for operational state and capabilities.

There are 14 principles that constitute how PowerFlex Manager was built. PowerFlex Manger is an easy to consume set of capabilities, a tool that users login through a UI or has API access. Its focused for operational purposes and for generalist to do all of their tasks in a single pain of glass. It also has the depth to execute common tasks that may be performed by a generalist or specialist, like lifecycle management, serviceability workflows, and expansion workflows. In totality, the goal is to balance the complex needs of the IT organization and ease of use.

The goal of PowerFlex Manager is not to replicate the capabilities of the subcomponents of the solution, that will be done in the element manager, such as the PowerFlex user interface, to turn such things on and off. However, if PowerFlex Manager needs to interface with PowerFlex capabilities through a broader workflow that may require a snapshot to be turned on, PowerFlex Manager can do this through an API call.

To explain the basics of how PowerFlex Manager works, I like to compare it to baking a cookie.

Through the use of Resources, Template and Services, PowerFlex Manager is able to Discover, Deploy and Operate PowerFlex systems.

The Resources are the servers, switches and hypervisor managers (such as Vcenter), which are physically added to a system and then “discovered” by PowerFlex Manager…if we were baking, these would be the ingredients. The template is essentially the recipe…templates are the requirements for settings and characteristics to be applied to a Resource or group of Resources while provisioning a Service.  These include the firmware, software, OS/Hypervisor, etc.  These requirements provide the input to be used in deployment automation. The outcome of all of this is a cookie, or what we call the Service, which is a group of Resources managed within PowerFlex as a Hypervisor Cluster.  These Resources are managed by PowerFlex Manager to provide health monitoring, compliance checking and remediation, and automated addition or removal of Resources as needed.

The PowerFlex Manager tool uses IT Operations Management (ITOM) Principles for its framework.  The goal of PowerFlex Manager is to enable rapid operational response to customers’ ongoing and ever-changing business needs, which we do through six key areas.

Start at 12 o’clock and move clockwise.

Deploy

There are some pre-requisites when deploying with PowerFlex Manager, such as the PowerFlex architecture. Rack n stack must be complete, PowerFlex Manager requires ethernet connectivity, and initial IP address. Lastly, PowerFlex Manager needs to be installed.

Once these pre-requisites are complete, PowerFlex Manager can deploy Services. Services are deployed through Templates. Templates are aligned to the offer, with Hyperconverged (HC), Compute Only (CO) and Storage Only (SO) to allow users to take the Resources available and deploy them into Services.

The deployment will take some time, the Template will take all of the Resources and balances it for deployment. For example, to deploy a 10 node cluster, it will take about 3 hours to complete all of the capabilities, then it is ready to move into operational state. This is a one time function, following it is them in the operational state of IT Operations Management (ITOM).

Health / Alerts

The second key area PowerFlex Manager focuses on is health and alerting. Within PowerFlex Manager you can see the outcome of a deployed system, it automatically turns on the interfaces to capture alerts and logging information flows through PowerFlex Manager. This gives the user health status at a glance, as depicted by the green check marks throughout this example. However, if something is setup incorrectly or not configured, it will also be represented in the UI.

PowerFlex Manager has the ability to dynamically show health and then pass along alerts via Secure Remote Services (SRS) or local NMS. Customers can configure the alert connector to send email alerts, in addition to, or instead of SRS alerts. PowerFlex Manager supports both SNMP and Syslog forwarding.

Compliance

The next key area is compliance. On an ongoing basis, PowerFlex Manager uses either the RCM (Release Certification Matrix for PowerFlex rack) or the IC (Intelligent Catalog for appliance) which defines how the system should be configured. This tracks the types of Resources and versions by number. PowerFlex then compares to determine if there has been any drift away from the defined version and flags it in the system to identify if in or out of compliance.

From the Resource pane, a user can easily export a compliance report to PDF or CSV.

Maintain


Another key area PowerFlex Manager covers is maintenance. Users are able to maintain their system using PowerFlex Manager maintenance modes.

One common example is a drive failure over time, a user can easily replace a drive using the Drive Replacement serviceability function, this can be done directly from the UI. PowerFlex Manager will prompt the user through a series of screens to replace the drive while PowerFlex Manager resets and rebuilds it into the system.

Automated Service/Drive Replacement Mode process overview


This process overview of PowerFlex Manager servicing options illustrates the simplicity of operations. With the RTS of PowerFlex Manger 3.2, there are now two options, Service Mode OR Drive Replacement can be used when performing services on nodes or executing drive replacement.

A single click is used to Put the Storage Data Server into Maintenance Mode in PowerFlex

Power down SVM And put ESXi host in maintenance mode in vSphere

Technicians can then perform manual maintenance, when completed the user clicks “Exit Service Mode/Exit Drive Replacement” and the automated operations continue.

The server is booted to ESXi

Take ESXi host out of maintenance mode

The SVM is powered on

And finally the SDS is taken out of maintenance mode.

As an example – if a fan goes out on a node, the error would be reported to Dell EMC support. Working with our customers the field tech replacing the fan can gracefully move that node into Service mode, replace the fan and then re-activate it back into the cluster.

Expand


PowerFlex Manager has built-in expansion capabilities. When starting small, PowerFlex Manager uses the concept of duplication, allowing users to expand by duplicating existing Services, matching the configuration. Users are able to expand clusters using the same tool used to deploy and operate.

Once the physical prerequisites are met, nodes are rack/stacked/cabled and given an initial IP address, PowerFlex Manager is them able to take that Resource and expand the logical Service onto it.

Upgrade


Lastly, PowerFlex Manager can be used for upgrading. PowerFlex ties together compliance and upgrade, so when changing between major version of the RCM/IC, PowerFlex Manager will show you all the difference associated with that version change. PowerFlex Manager now also supports all the large components of upgrades. Initially PowerFlex Manager started with BIOs, firmware and drivers, then added NXOS and ESXi, and now also supports PowerFlex and CloudLink update capability within PowerFlex Manager.

Once the upgrade is applied it does it in a rolling fashion, however a maintenance window is still recommended, but you are then able to upgrade your whole system.

PowerFlex Manager Getting Started screen


When you first start using PowerFlex Manager the “Getting Started” button located in the upper right of the interface takes you to a how to walk through for different tasks including managing firmware and compliance, defining networks, discovering resources, and working with services and templates.

PowerFlex Manager dashboard


No matter what operational state, everything is reflected in the PowerFlex Manager dashboard. The user design group has tried to display PowerFlex Manager in a way that summarizes all of the information in an easy to consume dashboard in a single glance.

On the underlying hardware, data is collected in 1 minute increments and showed in a 5 minute rollup. The data can be viewed in 5 minute, hourly, daily, week and monthly interval for up to one year.

PowerFlex Manager Resources


Here is the PowerFlex Manager Resources page showing a list of all managed resources for a Rack. Again, note that PowerFlex Manager includes the Cisco Nexus switches management in addition to our storage, compute, or HCI nodes.

What’s new in PowerFlex Manager 3.5

Better Management & Operations with Expanded Platform &
Device Support
Supporting the Install Base
New
Alerting & Mgmt Capabilities
New
Support for PowerFlex 3.5
New
Lifecycle Mode Improvements
  • Email alerting – configure alert connector to send email alerts
  • Manage volumes – ability to grow existing volume
  • Support for best practices when deploying
  • Async Replication (automate configuration of storage nodes to be ready for replication, inclusive of Static Routes and L3 networking connectivity)
  • Support for Protected Maintenance Mode (PowerFlex Manger service mode and upgrades now allows for IMM or PMM)
  • Support for new logical networking best practices v3
  • Enhancements to Lifecycle Mode

– Expanded use cases for additional HW configurations

– GUI improvements

– Service Mode capabilities extended to Lifecycle Mode (inclusive of Fault Sets)

  • Support for TACACS

Alerting & Management

  • You can configure the alert connector to send email alerts to one or more recipient email addresses you provide. You can receive these alerts in addition to, or instead of, Secure Remote Services alerts.
  • After adding volumes to a service, you can now resize these volumes. This feature applies to all types of services. For a storage-only service, you can increase the volume size. For a compute-only service, you can increase the size of the datastore associated with the volume. For a hyperconverged service, you can increase the size of both the volume and the datastore.

PowerFlex 3.5

  • When you deploy a service, PowerFlex Manager ensures that new PowerFlex 3.5 configurations use best practices for all components. PowerFlex allocates appropriate CPU, memory, and storage resources for the SVMs, PowerFlex Gateway, and presentation server. PowerFlex Manager also configures the networking to suit the requirements of these components.
  • You can deploy storage-only services with replication enabled. Replication is a feature of PowerFlex 3.5 that allows you to mirror the data across different geographical sites using native volume-level asynchronous replication. PowerFlex Manager deploys and configures the Storage Data Replicator (SDR) on all SDS nodes. PowerFlex Manager configures the journal capacity before adding the SDR. To facilitate replication, PowerFlex Manager has introduced a new replication network type
  • When you put nodes in service mode, or perform updates to the nodes or to the PowerFlex Gateway, you can take advantage of the Protected Maintenance Mode (PMM) feature that has been added to PowerFlex 3.5. Protected Maintenance Mode allows you to perform maintenance or updates that require longer than 30 minutes in a safe and protected manner. When you use Protected Maintenance Mode, nodes are temporarily removed from the protection domain while PowerFlex makes a temporary copy of the data
  • PowerFlex Manager supports logical network v3 configurations. This support covers new deployments, and upgrades and expansions on previous configurations. The sample templates have also been updated to include appropriate fields for logical network v3 settings.

Supporting the Install Base

  • Lifecycle Mode allows a service to perform only monitoring and compliance upgrade operations. All other service operations are blocked when a service is in Lifecycle Mode. Lifecycle Mode controls the operations that can be performed for configurations that have limited support. When you add a service, PowerFlex Manager automatically puts the service in Lifecycle Mode if the configuration has limited support.
  • PowerFlex Manager supports the ability to put multiple nodes in a single fault set into service mode when you must perform maintenance on those nodes. PowerFlex Manager supports no more than eight nodes in a single fault set as long as they belong to the same fault set. You can select nodes across the fault sets to put into service mode. You can choose to put the nodes into instant or protected maintenance mode. You can put a single node or all of the nodes in the same fault set into protected maintenance mode. If the Storage Data Replicator (SDR) is enabled, you can put only a single node into maintenance mode.

Lifecycle Mode Improvements


In addition to supporting PowerFlex 3.5 (Replication and PMM), PowerFlex Manager made a few enhancements to Lifecycle Mode.

In the last release we introduced the idea of Lifecycle Mode, which has really come to life in this release. This widens the scope of those who can take advantage of PowerFlex Manager. ​In this release we further expanded the guardrails for Lifecycle Mode, adding support for additional HW configurations. We also now show the user which Mode each Service is running in within the UI. And we added Service Mode capabilities within the scope of Lifecycle Mode.

Below, you can see a shot demo, exploring PowerFlex Manager

and below, you can see a demo, which shows how to Deploy Service and Create Volumes using Dell PowerFlex Manager

Leave a Reply