|Though I’d rather be writing this blog over a coffee and beignet in New Orleans, I’m still thrilled to be sharing the news about our HPE Storage Solutions for Microsoft, and specifically what’s new for the Microsoft Ignite audience. As we all know, this year’s Microsoft Ignite is taking place on the Web rather than the Crescent City, but our Solution team is still treating it like it’s our annual “Mardi Gras”. With that in mind, here’s what new for our HPE Storage Solutions for Microsoft. Let the good times roll! |
Throw me somethin’, Mista!
HPE is an official sponsor of this year’s Microsoft Ignite show, which runs September 22-24. Along with the presence within the online venue, HPE has created its own web presence to supplement things, our “Virtual Booth”, due to all the related activities and content we had available. This pandemic has accelerated digital transformation across the company’s customer base, and in response HPE is stepping up the number of solutions to help these organizations achieve new ways of working and serving their end-customers, especially in the cloud.
A key part of HPE-at-Ignite presence will be a series of Expert Videos. Storage team experts will deliver a few of these sessions, coverings topics from new a SQL Server solution built on a newly re-engineered storage platform, to Hybrid cloud data expansion, to Big Data HCI.
The HPE Microsoft Ignite 2020 Virtual Booth landing page is the place to watch these video sessions, as well as get the latest Microsoft Storage Solution resources including technical whitepapers, solution briefs and other online assets.
Starting with Microsoft Ignite week and beyond, we’re also rolling out more expert video content that couldn’t fit on the Landing page. These sessions are being delivered through the HPE Storage BrightTalk channel and are spanning topics from SQL Server Big Data Clusters, to Storage Class Memory, Highly Available Windows File Services, Hybrid Cloud data mobility, and Microsoft VDI. These sessions are supported by our strategic partner, Intel.
Who Dat HPE MSA Gen 6?
The company recently broke the news on the newly updated HPE Modular Smart Array (MSA) storage, which features a new architecture, ASIC, chipsets and health monitoring capabilities. The product team shared the key details in that HPE MSA Gen6 Announcement, for this storage product that is reaching new levels of ease-of-use plus price/performance. I’d like to highlight what it means for our Microsoft storage solutions, which I covered in more detail in the SQL Server on HPE MSA solution release blog.
The higher levels of performance from the Gen6 platform will translate directly into more transactions for SQL Server databases, more apps hosted across your Hyper-V environment, and even greater Microsoft workload consolidation onto a single array. The new Tiering 2.0 algorithm alone delivers up to 45% more app acceleration than in the previous generation system.
Other enhancements will increase workload availability, such as MSA-DP+, an advanced RAID-based data protection feature which protects data and enables faster rebuilds, and the Health Check tool which makes it easy to help ensure optimal system operations.
HPE MSA storage has traditionally been used for entry-level and departmental SQL Server database environments. We expect this to continue, especially where organizations require on-premises control of their data and the hands-on ability to ensure services levels. There are also new realizations shared on a Solution team member’s blog regarding the benefits of a single RAID-protected Storage array versus having to maintain multi-node clusters of HCI systems for the same workload and data capacity.
Innovation that takes the (King) cake
The solution development hasn’t stopped at HPE since the last Microsoft Ignite event. The Storage krewe has led a parade of new offerings to meet the needs of our broad Microsoft workload customer base:
Nimble Storage Extender for Azure Stack Hub – Need more data capacity for your Azure Stack Hub but don’t want to buy a whole new one? The HPE Storage Extender for Azure Stack Hub brings flexibility to the tightly defined Azure Stack Hub architecture, letting you just expand data capacity plus get the benefits of enterprise SAN data services.
Windows Admin Center (WAC) Extensions – We’ve rolled out a number of WAC Extensions to expand the visibility and manageability of our storage products within the WAC dashboard. These include Windows and Azure Stack HCI extensions for the HPE Apollo 4200, as well as Storage Extensions for the HPE Primera and HPE 3PAR storage platforms.
HPE InfoSight for Hyper-V – Breakthrough new HPE Storage AI capability that brings cross-stack analytics parity for Windows Hyper-V VMs that was previously only available for VMware ESX environments.
Accelerated SQL Server on HPE Nimble Storage – Takes new NVMe SSD caching approach to increase performance for OLTP and other demanding apps. Lab verified benefits of Storage Class Memory powered by Intel Optane SSDs show back-end latency reduced as much as 50%. Similar lab validated solution also available for HPE 3PAR storage.
SharePoint 2019 & Skype for Business Server 2019 on HPE Storage – Upgrade your SharePoint and Skype environments on-premises to the latest Server 2019 versions to modernize your infrastructure and take advantage of the latest in content collaboration and portal technologies.
SQL Server Big Data Clusters (BDC) – A single scale-out solution for both relational and Big Data, built on HPE Storage and the latest data center technologies including containers and Kubernetes. Manage more data from across the enterprise with your existing SQL Server tools and expertise.
In addition to these HPE Storage Solutions for Microsoft releases over the past year, there are even more in progress, including expanded testing and technical publications for the SQL Server BDC PolyBase feature, SQL Server leveraging HPE Primera storage enhancements, and SQL Server realizing new levels of performance with persistent memory (PMEM).
Get (the party) started
So if you haven’t already, definitely get registered for Microsoft Ignite. Then mix yourself a Hurricane or Sazerac, and go straight to the HPE Microsoft Ignite event Virtual Booth site to join the party with all our event activities, on-demand content and technical resources.
And don’t forget our Storage news of the show, how the new HPE MSA Gen6 Storage can be the life of your party for Microsoft workloads – with more application performance, yet simplified management that fits any IT budget. Learn more at: the SQL Server 2019 on the new HPE MSA Gen6 solution Blog
The HPE Discover Virtual Experience starts Tuesday June 23rd, when 10s of thousands of people will join online to learn about new technologies to transform their businesses such as intelligent edge, hybrid cloud, IoT, Exascale computing and much more. Our team will be part of this event, showing off our newest solution for Microsoft SQL Server Big Data Clusters running in the HPE Container Platform. Here’s the direct link for session and speaker information or search us up once you’ve registered and join the event — we’re session “D139“.
The inspiration for our work is that data growth is taking off like a rocket, and in that spirit the HPE Storage team staged our approach to the new enterprise database capability from Microsoft: SQL Server 2019 Big Data Clusters. We lifted off with an initial enterprise-grade solution for SQL Server Big Data Clusters (BDC), and laid-in a course for more features, capabilities and scale. As introduced in previous blogs, SQL Server BDC uses a new architecture that combines the SQL Server database engine, Spark and Hadoop Distributed File System (HDFS) into a unified data platform.
This approach escapes the gravitational constraints of traditional relational databases, now having the ability to read, write, and process big data from traditional SQL or Spark engines, letting organizations combine and analyze high-value relational data along with high-volume big data, all within their familiar SQL Server environment. Our first stage effort includes an initial implementation guide, collateral and a number of related activities including a live demo in this year’s HPE Discover Virtual Experience.
Following soon will be ‘stage 2’ where we’ll publish technical guidance on deploying your own BDC that takes advantage of data virtualization, also known as the Polybase feature. Polybase lets you virtualize and query other data sources from within SQL Server without having to copy and convert that outside data. It eliminates the time and expense of traditional extract, transform, and load (ETL) cycles, and perhaps more importantly, lets organizations leverage existing SQL Server expertise and tools to extract the value of third-party data sources from across the organizational data estate such as NoSQL, Oracle, and HDFS, to name just a few.
The last stage of this mission will add HPE Apollo 4200 storage systems for a cost-effective storage pool, especially for larger BDC deployments in the petabytes.
Info on our overall SQL Server BDC solution is available online in the new solution brief.
Putting BDC boots on the moon
There are a number of key considerations for deploying your own SQL Server BDC. It’s going to be a very different environment than what you may be familiar with for traditional Microsoft SQL Server. Rather than a Windows environment, with or without VMs, BDC requires the use of containers and along with running on Linux, the architecture will contain a number of possibly new technologies for traditional IT teams: Kubernetes, Apache Spark, Hadoop Distributed File System (HDFS), Kibana and Grafana.
Many companies have begun to use Kubernetes as an efficient way to deploy and scale applications. It’s often referenced as a key part of a typical Continuous Integration and Continuous Deployment (CI/CD) process. And one survey puts the number at 78% of respondents using Kubernetes in production. So bringing Kubernetes to SQL Server may be a timely way to merge a couple areas of significant investment for companies: traditional RDBMS and the evolving DevOps space.
Another unique feature of this solution is Container management. Our initial technical guidance includes the use of the HPE Container Platform. The HPE Container Platform provides a multi-tenant, multi-cluster management infrastructure for Kubernetes (K8s). Creating a highly available K8s cluster is as easy as importing the hosts into the platform and defining master/worker role. In addition, it simplifies persistent access to data with the integration of Container Storage Interface (CSI) drivers. This makes connecting with HPE storage easy, not only providing persistent volumes, but enabling access to valuable array-based resources such as encryption and data protection features like snapshots. The latest HPE CSI package supports HPE Primera storage, HPE Nimble storage and HPE 3PAR storage.
Key components of the initial solution include:
- Microsoft SQL Server 2019 Big Data Clusters
- HPE ProLiant DL380 Gen10 servers
- CentOS Linux—a community-driven, open source Linux distribution
- HPE Nimble Storage arrays for the master instance to provide integrated persistent storage
- HPE Container Storage Interface (CSI) driver
- Kubernetes to automate deployment, scaling, and operations of containers across clusters of hosts
- HPE Container Platform for the deployment and management for Kubernetes clusters (optional)
- HPE MapR as an integrated, persistent data store (optional)
Why HPE Storage for Big Data Clusters
The partnership of Microsoft and HPE stretches back to the same time that the Hubble space telescope was launched, about 30 years ago. This heritage of testing and co-development has helped ensure optimal performance for Microsoft business software on HPE hardware. Other important reasons to chose HPE for your BDC deployment:
- HPE developed a standards-compliant CSI driver for Kubernetes to simplify storage integration.
- HPE developed the HPE Container platform, providing the most advanced and secure Kubernetes-compatible container platform on the market.
- HPE owns MapR, an established leading technology for big data management — now incorporated within the HPE Data Fabric offering — and another key part of the solution that helps span data management from on-premises to the cloud
- Finally, HPE has had in the market a complete continuum of SQL Server solutions based on HPE Storage – from departmental databases to consolidated application environments, and from storage class memory accelerated to the most mission-critical scale-up databases. Adding BDC provides yet another option – now for scale-out data lakes – to customers who rely on HPE as a trusted end-to-end solution partner.
The HPE Storage with Microsoft SQL Server Big Data Clusters solution is available today. An initial reference architecture delivers the benefits of scale-out SQL Server on HPE Nimble enterprise-class data storage with the newest container management capability using the HPE Container Platform.
The HPE Storage with Microsoft SQL Server Big Data Clusters solution is a safe, first step for your IT team, but a giant leap forward for your organization to derive the most business value from its data estate, regardless of whether its relational, unstructured, on-premises or in the cloud.
Learn more about HPE Storage solutions for Microsoft and see us live at the HPE Virtual Discover Experience.
Are you struggling to manage more data, and more types of data from across the enterprise? Start your mission to manage your entire data estate with existing SQL Server expertise. Read the new implementation guide: How to deploy Microsoft SQL Server 2019 Big Data Clusters on Kubernetes and HPE Nimble Storage.
In early 2016 Microsoft announced their Microsoft Azure Stack Hub, along with their new Windows Server 2016, as an extension of the Microsoft Azure Public Cloud. From the beginning, customers sought the benefits of Azure cloud, but deployed within their own datacenter. This new hybrid cloud solution brought them the hyperscale experience of consumption-based computing, along with the ability to run the very same applications you could run on Azure.
This on-premises version of the cloud came with constraints, however. Not all of the same services in Azure are available in Azure Stack. It’s not ‘limitless’ as it is bounded by the capacity and compute contained within a particular customers’ Azure Stack deployment. And that configuration itself is rigidly defined by Microsoft, and available only through a few certified partners, selling predefined bundles of servers, storage and networking.
Up until now, even if they didn’t need more compute or bandwidth, if a customer’s storage needs exceeded that of the configuration that they purchased, they would have to purchase another node (predefined set of compute/storage/networking) or an entire new Azure Stack offering. With Azure Stack, the promise of cloud agility came with strict boundaries.
But that has all changed.
Expand Hybrid cloud capacity with the HPE Storage Extender
The new HPE Storage Extender for Azure Stack Hub solution leverages and delivers scripts and technical guidance on how to expand just the data storage capacity of your Azure Stack Hub environment. This optimized implementation is initially available on HPE Nimble Storage, and works in parallel with Microsoft’s published approach for expanding capacity for Azure Stack Hub by using any iSCSI storage resource.
HPE Nimble Storage brings intelligent, self-managing flash storage to your data center and hybrid cloud environments. It is an ideal platform for your expanded Azure Stack Hub, with high availability and Data Efficiency features, and guarantees both data availability of 99.9999%1 as well as the ability to store more data per terabyte of flash storage than other all-flash arrays. Designed for advanced storage technologies such as NVMe, HPE Nimble Storage delivers industry-leading capacity efficiency as well as a future-proof architecture. You can find guaranteed data-reduction program details in the Store More Guarantee documentation2.
Azure Stack Hub and Data storage
Azure Stack Hub scales between 4 and 16 hybrid nodes and 4 and 8 all-flash nodes. Although the Azure Stack Hub does not contain all of the features of Azure public cloud, it mimics the most common features, and provides an easy transition point for data and applications moving between the cloud and on-premises. The primary method of managing an Azure Stack Hub instance is through an almost identical management portal of the Azure public cloud.
Microsoft Azure Stack Hub currently provisions storage utilizing internal disk from hyperconverged nodes managed by Storage Spaces Direct (S2D). Up to this point, external storage has not been supported under the Microsoft Azure Stack Hub design options; the total capacity and performance available was capped by the maximum number of nodes in the scale unit, the disk drive configurations available from each OEM vendor, and the specific characteristics of the virtual machine type deployed.
The tightly enforced configurations have been at odds with customers and partners requests for flexibility, and specifically for the ability to leverage external storage arrays to support key workloads. Along with additional data capacity, external storage arrays bring with them the ability to migrate and replicate data, along with higher data availability. This is why HPE developed a means to connect HPE Nimble Storage arrays as an external iSCSI storage option, in parallel with Microsoft’s technical template for how to connect to iSCSI storage with Azure Stack Hub3.
HPE brings innovation to Windows with the world’s most intelligent storage
The HPE Storage Extender for Azure Stack Hub solution provides the key elements to enable access to external data capacity, while maintaining a customer’s supported Azure Stack Hub configuration. The solution includes:
- HPE Nimble Storage, officially Windows Server 2019 and 2016 certified
- Windows Server 2016 Datacenter or Windows Server 2019 Datacenter (latest build recommended)
- PowerShell DSC extension
- Custom Script Extension
- Solution Deployment Guide
- HPE InfoSight
HPE InfoSight — the IT industry’s most established AI platform — is the key feature in enabling autonomous, self-managing data storage, and is an embedded part of HPE Nimble Storage, as well as other HPE Storage and Server products. HPE InfoSight has analyzed application patterns in 1,250 trillion data points over the last decade to predict and prevent disruptions across storage, servers, and virtual machines. This has resulted in savings of more than 1.5 million hours of lost productivity due to downtime. HPE InfoSight provides the intelligent foundation for all HPE storage products, creating the industry’s only end-to-end AI capability for self-managing storage.
The HPE Storage Extender for Azure Stack Hub solution is available today. It brings additional data capacity to Azure Stack Hub — without the cost of adding additional compute.
Thanks to HPE Nimble Storage, the HPE Storage Extender for Azure Stack Hub solution is an economical way to access and use more data within your Microsoft Hybrid cloud, while benefitting from improved data management, protection and availability. The solution includes technical guidance and scripts, and is supported by Microsoft as an approach aligned with Microsoft published technical templates.
Solution resources available at launch:
Learn more about HPE storage solutions for Microsoft @ hpe.com/storage/microsoft
Microsoft Azure Stack HCI has established itself as a solid hyperconverged infrastructure offering, based on the leading operating system, Microsoft Windows Server 2019. IT staff are able to efficiently consolidate traditional workloads on this familiar platform, thanks to multiple technological features including both compute virtualization with Hyper-V as well as data storage virtualization with Storage Spaces Direct. There’s also support for the use of non-volatile memory express (NVMe) SSDs and persistent memory for caching in order to speed system performance.
However, with such dynamic technology in play at the OS layer, things get interesting when you add a sophisticated workload that also has its own intelligent performance enhancing features including storage tiering, a metacache database (MCDB), and dynamic cache. In this case we’re talking about Microsoft Exchange email, which recently introduced the new Microsoft Exchange Server 2019.
One Wall Street firm was a power user of Microsoft Exchange – with over 200,000 users, many having massive mailboxes of dozens up to 100 or more GBs in size. As part of their infrastructure planning, the customer wanted to compare the performance and cost of continuing to run Exchange on physical servers with external attached storage (JBOD), versus evolving to an Azure Stack HCI infrastructure.
The combination of these products and technologies required complex testing and sizing that pushed the bounds of available knowledge at the time, generating learning useful for other companies who are also early in adopting various combinations of demanding enterprise workloads on top of Azure Stack HCI.
Field experts share their insight
“This customer had an interest in deploying truly enterprise-scale Exchange, and eventually the latest server version, using their HCI infrastructure,” began Gary Ketchum, Sr. System Engineer in the Storage Technology Center at HPE. “Like vSAN or any other software-defined datacenter product, choosing the hardware is very important in order to consistently achieve your technical objectives.”
This observation especially holds true when implementing Storage Spaces Direct solutions. As stated in the Microsoft Storage Spaces direct Hardware requirements page, “Systems, components, devices, and drivers must be Windows Server Certified per the Windows Server Catalog. In addition, we recommend that servers, drives, host bus adapters, and network adapters have the Software-Defined Data Center (SDDC) Standard and/or Software-Defined Data Center (SDDC) Premium additional qualifications (AQs). There are over 1,000 components with the SDDC AQs.”
A key challenge of the implementation was in how to realize the targeted levels of improved flexibility, performance, and availability, within a much more complex stack of technologies, multiple virtualization layers, including potentially competing caching mechanisms.
Anthony Ciampa, Hybrid IT Solution Architect from HPE explains key functionality of the solution. “Storage Spaces Direct allows organizing physical disks into storage pools. The pool can easily be expanded by adding disks. The Virtual Machine VHDx volumes are created from the pool capacity providing fault tolerance, scalability, and performance. The resiliency enables continuous availability protecting against hardware problems. The types of resiliency are dependent on the number of nodes in the cluster. The solution testing used a two-node cluster with two-way mirroring. With three or more servers it is recommended to use three-way mirroring for higher fault tolerance and increased performance.” HPE has published a technical whitepaper on Exchange Server 2019 on HPE Apollo Gen10 available today online.
Microsoft Azure Stack HCI on HPE Apollo 4200 Gen10 solution
At Microsoft Ignite 2019, HPE launched its solution for the new Microsoft HCI product, Windows Azure Stack HCI with HPE Apollo 4200 Gen10. This new software-defined hyperconverged offering, built on the high capacity yet dense Apollo storage server, delivered a new way to meet the needs of the emerging ‘Big Data HCI’ customer. A new deployment guide details solution components, installation, management and related best practices.
The new Azure Stack HCI on HPE Apollo 4200 solution combines Microsoft Windows Server 2019 hyper-converged technology with the leading storage capacity/density data platform in its class. It serves a growing class of customers who want the benefits of a simpler on-premises infrastructure while still able to run the most demanding Windows analytics and data-centric workloads.
Findings from the field
Notes from the deployment team captured some of the top findings of this Exchange on Windows HCI testing, that will help others avoid problems as well as confidently speed these complex implementations.
- More memory not required – The stated guidance for Azure Stack HCI requires additional memory, specifically an SSD NVMe (cache tier) beyond JBOD physical deployment. However HPE’s Jetstress testing showed that similar performance was also possible from just JBOD. Thus the server hardware requirements are similar between Azure Stack HCI and JBOD, and even if the customer plans to deploy JBOD MCDB tier with Exchange 2019, the hardware requirements are still very similar. Note, there could be other cost factors to consider such as the cost of overhead for additional Compute and RAM within the Azure Stack HCI, as well as any other additional software licensing cost for running Azure Stack HCI.
- Size cache ahead of data growth – The cache should be sized to accommodate the working set (the data being actively read or written at any given time) of your applications and workloads. If the active working set exceeds the size of the cache, or if the active working set drifts too quickly, read cache misses will increase and writes will need to be de-staged more aggressively, hurting overall performance.
- More volumes the better – Volumes in Storage Spaces Direct provide resiliency to protect against hardware problems. Microsoft recommends the number of volumes is a multiple of the number of servers in your cluster. For example, if you have 4 servers, you will experience more consistent performance with 4 total volumes than with 3 or 5. However, testing showed that Jetstress provided better performance with 8 volumes per server compared to 1 or 2 volumes per server.
Where to get more info
Microsoft Azure Stack HCI on HPE Apollo 4200 Gen10 server is a new solution that addresses the growing needs of the Big Data HCI customer – those who are looking for an easy-to-deploy and affordable IT infrastructure with the right balance of capacity, density, performance, and security. Early work with this solution, especially where it’s being combined with demanding and data intensive workloads, can create non-intuitive configuration requirements, so IT teams should seek out experienced vendors and service partners.
A new deployment guide details solution components, installation, management and related best practices. Information in that document, along with this blog, and future sizing tools expected out from HPE, will continue to provide guidance for enterprise deployments of this new HCI offering.
The deployment guide is available online today at this link: <link to Deployment Guide>
If you were at Microsoft Ignite last month you may still have missed the launch of HPE’s latest hyperconverged infrastructure (HCI) solution: Microsoft Azure Stack HCI on HPE Apollo 4200 storage. It would be understandable, as Ignite was a major industry event packed with technology news, especially with lots of HPE show activity, including prominent HPE mainstage appearances for both Azure Stack and the new Azure Arc.
But among the new and enhanced solutions we demonstrated at the show, our presentations about Azure Stack HCI on HPE Apollo storage were well-received and timely given the growing emphasis on HCI, hybrid cloud and all things software-defined. The key message for this solution was that it is pioneering a new area in software-defined HCI for Windows Big Data workloads. The solution uniquely delivers the convenience of hyperconverged Infrastructure on a high-capacity platform for the most data-intensive applications.
The emergence of Big Data HCI
We’ve all heard about the explosive growth of data, and that we’re in an age of zettabytes. IDC made a specific prediction, that by 2024, just data created from AI, IoT and smart devices will exceed 110 zettabytes (source: IDC FutureScape: Worldwide Cloud Predictions 2020).
At the same time, organizations are trying to simplify their IT infrastructures to reduce cost, complexity and the need for specialized expertise. The conflict is that the applications required to harvest this explosion of data can be the most demanding in terms of performance and management. I’m seeing companies – even the largest most capable enterprises – are recognizing the value of easy-to-use hyperconverged infrastructure to alleviate some of the strain of delivering these demanding, data-centric workloads.
Azure Stack HCI on HPE Apollo 4200 storage is a new solution that addresses the needs of the growing “Big Data HCI” customer. Azure Stack HCI on HPE Apollo is built on the highest capacity Azure Stack HCI qualified 2U server, bringing an unmatched ability to serve big data workloads on a compact Windows software-defined HCI appliance.
HPE Apollo HCI solution key components
Azure Stack HCI is Microsoft’s software-defined HCI solution that pairs Windows Server 2019, Hyper-V, Storage Spaces Direct, and Windows Admin Center management, along with partner x86 hardware. It is used to run Windows and Linux VMs on-premises and at the edge with existing IT skills and tools.
Azure Stack HCI is a convenient way to realize benefits of Hybrid IT, because it makes it easy to leverage the cloud-based capabilities of the Microsoft Azure cloud. These cloud-based data services include: Azure Site Recovery, Azure Monitor, Cloud Witness, Azure Backup, Azure Update Management, Azure Network Adapter, and Azure Security Center to name a few.
The Azure Stack HCI solution program includes Microsoft-led validation for hardware, which ensures optimal performance and reliability for the solution. This testing extends to technologies such as NVMe drives, persistent memory, and remote-direct memory access (RDMA) networking. Customers are directed to use only Microsoft-validated hardware systems when deploying their Azure Stack HCI production environments.
HPE Apollo 4200 Gen10 Server – leading capacity/throughput for Windows HCI
The HPE Apollo 4200 Gen10 server delivers leading scale and throughput for Azure Stack HCI. The HPE Apollo 4200 storage system can accommodate 392 TBs of data capacity within just a 2U form-factor. This leads all other Azure Stack HCI validated 2U solutions as seen in the Microsoft Azure Stack HCI catalog (Microsoft.com/HCI). In addition, the HPE Apollo storage system is a leader in bandwidth, supporting 100Gb Ethernet and 200Gb Infiniband options. Customers are already running large scale, data-centric applications such as Microsoft Exchange on HPE Apollo systems, and can now add Azure Stack HCI as a means to simplify the infrastructure stack, while preserving performance and the space-efficient 2U footprint.
The HPE Apollo Gen10 system is future-proofed with Intel Cascade lake processors for more cores and faster processing, along with memory enhancements and support for NVMe storage. The HPE Apollo systems leverage a big data and high performance computing heritage, and have an established Global 500 customer track record.
Azure Stack HCI on HPE Apollo solution – more than just hardware
The HPE Apollo 4200 system is at the core of this Microsoft software-defined HCI solution, but there’s much more to the solution. HPE solution engineering teams perform testing on all solution designs, and publish technical whitepapers to provide guidance on implementation, administration, and performance optimization, for example the recent Microsoft Windows Server 2019 on HPE Apollo 4200 implementation guide. HPE also trains authorized reseller partners to help assure fast, successful deployments and fast time-to-solution for customers.
Windows Admin Center (WAC) is becoming the new standard interface for Windows system management. HPE is developing Extensions for WAC that will make it easier to manage HPE Apollo systems within Windows Server 2019 environments as well as specifically within Azure Stack HCI clusters.
As an HPE Storage solution, customers also enjoy high availability through HPE InfoSight predictive analytics that deliver the uptime benefits of AI to the datacenter.
Get started with HPE Apollo HCI
The Azure Stack HCI on HPE Apollo solution is available today. It’s the largest capacity 2U Azure Stack HCI validated solution available, and has been officially qualified for All-Flash, Hybrid SAS SSD, and NVMe providing options for affordable and high-performance data storage.
The Azure Stack HCI on HPE Apollo solution is the go-to choice for analytics and data-centric Windows workloads. Get easy to manage infrastructure with native Microsoft Windows administration. Available with the solution are published technical guidance including whitepapers and related resources, with WAC extensions on the way.
The launch webinar was recorded and is available on demand – watch it to learn more:
Our solution engineers are constantly configuring, testing, benchmarking and documenting our Microsoft solutions, providing vital guidance to our customers and resellers so that they can quickly and confidently deploy them. I had a moment to go back over some recent work we did on our Microsoft Exchange Solution, and I wanted to share some highlights that can be missed if you didn’t have the time to read through these detailed technical publications.
The widening orbit of Apollo
First off, the HPE Apollo 4200 Gen10 has been our go-to Storage server for some time. A key offering in the HPE Storage product line, it serves as the company’s lead platform for Software-defined storage solutions. It shares a heritage with the Apollo HPC compute systems (e.g. model 2000s, 6000s) around HPC storage and big-data usage. Storage use evolved to include object storage, scale-out file systems, backup, archive and other data-intensive workloads. Of course Microsoft Exchange is the grand-daddy of business email, a classic data-intensive application, esp. if you’re regularly sharing multi-MB sized powerpoints.
As we continue to build-out our Microsoft solution business, we’ve hit upon a powerful use scenario for Apollo as a leading platform for Azure Stack HCI. You’ll continue to see more news and innovation coming from us in this area in the coming months.
HPE Apollo 4200 Gen10 server with Exchange 2016 ESRP
Our team undertook testing based on Microsoft’s testing benchmark/certification known as the Exchange Solution Reviewed Program. The program combines a storage testing harness (Jetstress) with solution publishing guidelines, and has been used by Storage OEMs for over the past decade as a standard testing framework. Customers and resellers have been eager consumers of the publications as a way to compare results across vendors and to help design their Exchange storage architectures and deployments.
This specific testing used the HPE Apollo 4200 Gen10 storage system, which is a 2U server available either in a 24 large form-factor (LFF) drive configuration or a 48 small form-factor (SFF) drive configuration.
The remarkable feature of the 4200 is that the drive cages use two trays, with the second tray accessed by simply pulling the front tray forward on internal rails and can be done while the server is operational. This unique chassis design produces probably the highest storage density available in a 2U server in the industry.
The bigger brains of the Apollo 4200 Gen10
The HPE Apollo 4200 Gen10 contains 16 DIMM slots, which support up to 1TB RDIMM (registered) or 2TB of LRDIMM (Load reduced) memory. Individual DIMM capacity has doubled in this latest platform update, going from 32GB to 64GB RDIMMs and 64GB to 128GB LRDIMMs.
The CPUs have been upgraded to use next-gen Intel® Xeon® Scalable processors, the same ones designed for AI/Deep learning workloads, and future-proofed with support for Intel Optane PMEM. The upgraded CPU enables an increase from 24 cores to 28 cores and with the clock speed increasing from 2.1 to 2.3GHz.
There are 5 PCIe slots accessible from the rear of the unit. The 16x slots support 100Gbe Ethernet connections, and the 24x slots support up to 6 NVMe SSDs.
Get flexible with the Apollo
Another feature of the HPE Apollo 4200 Gen 10 is its flexibility. Multiple configurations are possible, depending on capacity and performance needs, making this a valuable platform for Exchange deployments of any size.
The rear drive cage alone can be re-configured in a number of ways:
- Five low-profile PCIe slots with two processors
- Or, Four LFF rear drive cages, along with the Five low-profile PCIe slots with two processors
- Or, Two SFF rear drive cages + two full-height half-length risers, and the Five low profile PCIe slots with two processors
- Or, Six SFF rear NVMe cages, along with the Five low-profile PCIe slots with two processors
The Ultimate Exchange Building Block
This HPE solution for Microsoft Exchange Server 2016 was designed using a building block approach, with a multi-copy database design using Exchange Database Availability Groups (DAGs). A DAG is a group of up to 16 mailbox servers that host a set of databases, providing automatic recovery from failures that affect databases or individual servers.
This solution used a single DAG in two building blocks, which are two servers in the primary site and two servers in the secondary site, to support 4,000 mailboxes per building block with a 25 GB capacity, and a messaging profile of 200 messages sent and received per user, per day. Using the building block approach, customers can scale their Exchange environment to a size that fits their needs.
The Microsoft Jetstress testing validated that the storage subsystem was capable of the IOPS needed to support the configured number of mailboxes as well as providing additional headroom for growth.
Where to get the rest of the story
The HPE Apollo 4200 Gen10 with Exchange 2016 ESRP document is available online today. Entitled, “HPE Apollo 4200 Gen10 Server 4000 mailbox resiliency Microsoft Exchange 2016 using 8 TB 7.2K large form factor drives”, it is based on Microsoft’s ESRP 4.0 testing framework. The technical whitepaper contains details on the tested performance results, configuration best practices, product photos, and includes extensive Jetstress output.
The document can be downloaded here.
I was going to write a blog on the new office space my group is moving into, and my thoughts on it, and realized someone had already written about it: https://www.executiveforum.com/office-space-2019/. It’s interesting the universality of opinions on modern corporate Open Office design. In our case the move was a bit more stark, as we were exiting a start-up environment of comradery over free food, intensity of a shared mission, tight collaboration across organizations, and even a garden with bee hives(!) – which has been reengineered to an open office design, with piped-in ambient water sounds, as part of a far flung global organization.
Granted, much has been written about open plan offices being better for collaborative work and that they may be better aligned with the habits of a younger work force. It’s certainly more space efficient – taking the traditional average space per worker ranging to 250 square feet, now with these high density designs, down to 100 sf or less.
My take-away? Clearly office design needs to map to business objectives the same way compensation programs, supply chain, financial strategy, and corp dev have to. Being within the vortex of this change makes it challenging to be completely objective, but being a participant provides the credibility of directly experiencing the pros and cons. I’m seeing the Open Plan approach should be applied primarily to realize more efficient collaboration for integrative work-teams, not in cases where the workgroup process requires quiet attention to detail, thoughtful strategic thinking, or significant private/sensitive communications, to name a few.
Customers, partners and members of the HPE Community converged on Las Vegas this past week for Discover 2019. It’s the company’s largest customer event which provides hands-on learning and training on HPE products as well as direct contact with members of the teams who make them.
One of the topics covered in the show was Hybrid cloud, and how to keep important data protected and available. New performance statistics were released from recent testing, with one of them relating to Microsoft SQL Server. This comes at a time when the company has released an update to Recovery Manager Central for SQL Server (RMC-S), which has been a popular means to ensure Application consistent database copies.
HPE Storage solutions for MS SQL Server
HPE has been investing to deliver outstanding levels of both Performance and Manageability for SQL Server. First off, HPE Data Storage makes SQL Server Faster. HPE has multiple lines of All-Flash storage arrays HPE 3PAR StoreServ, HPE Nimble Storage, and now the all new HPE Primera, that let you run Microsoft SQL Server in production at the highest levels of performance and availability. With HPE flash storage you get reliably fast performance: up to 3.8 million IOPS (I/Os per second) on 3PAR, and about 2 million IOPS on a cluster of Nimble storage all-flash arrays. And all at a consistent sub-millisecond latency.
Recently we added to this performance boost with the addition of Memory-Driven Flash. This is an all-flash array with an NVMe Storage Class Memory (SCM) cache. Recent testing with a SQL Server database on Memory-Driven Flash has shown a 59% decrease in latency (sec/read) compared to when SCM is disabled. Decreased latency translates to faster database query responses and a better user experience.
We’ve also invested to make SQL Server administration better. HPE delivers the most consistent availability with Intelligent Storage and HPE InfoSight, delivering 99.9999% guarenteed uptime. This means your data remains available, and in the rare event that there is an issue, the storage proactively alerts you to potential problems. HPE Storage comes with copy data management capabilities, and extends them across your hybrid cloud. It’s easy to create SQL-consistent snapshots or archival copies to Azure. HPE Storage arrays support data workflows and tasks that can manage live copies on and off-premises for DR, test/dev, reporting, analytics, patching, & upgrading. And HPE Storage enhances SQL Server DevOps as we are the only vendor to offer rich data services for both Windows and Linux Containers, and have interop with all the leading container platforms including Docker, Kubernetes, Mesos. HPE storage was the first to provide persistent volumes for containers.
How Recovery Manager Central improves SQL Server data protection
HPE Recovery Manager Central software, and specifically the RMC-S plug-in, allows SQL Server administrators to protect SQL Server instances and databases with application-consistent recovery points. The snapshots are managed on HPE 3PAR storage, and you can use RMC-S with either HPE StoreOnce, HPE Data Protector, Symantec NetBackup, or Symantec Backup Exec to protect and restore these snapshots.
Recent testing has shown that creating a database clone with RMC-S takes just 1 minute and 22 seconds. This is compared to the traditional and very manual process of creating a clone with SQL Server which can take 45 mins. The automated process of RMC vastly outperforms the manual process of creating new database, copying the schema, and then inserting and verifying data within every individual table. Compared to the traditional approach, RMC-S lets you copy a database in one step and 32 times faster.
Where to get more info
There’s much more information available on the advantages of HPE storage for Microsoft SQL Server, as well as for all the important Microsoft business workloads such as Windows Server, Exchange and newer solutions like Azure Stack and Azure Stack HCI. Here are some key resources below:
• Webinar: Get more from your Data with HPE Microsoft Storage solutions
• Reference Architecture: Microsoft SQL Server with HPE 3PAR Memory-Driven Flash
• Blog: Cut SQL Server Latency in Half with Memory-Driven Flash Storage
• HPE Storage Microsoft Solutions webpage: HPE.com/storage/microsoft
There’s a very wicked ’55 Chevy lookin’ for you
I was nostalgic when I read the story about VxBlock hitting 10. I had been working in Converged computing going back some time, and helped grow the FlexPod business. So to acknowledge this passage of time also recognizes just how old this space is, esp. in technology terms (i.e. ‘Internet dog years’). This product started as an entire freestanding business – the Acadia then VCE joint venture back in 2009. The article cites how EMC once had a full line of Vblocks – which evolved into VxRack, VxRail and VxFlex among other offerings – but today this whole business and converged product line has since been trimmed to one single model, the VxBlock 1000, because as they say themselves, “they’re outdated”.
File that under C.S.
The Converged market is in the throes of a metamorphic change. What was once the big story, popular Certified Reference and Integrated systems like VxBlock and FlexPod revolutionizing the datacenter, are now shrinking and in decline, about a third of the market and dropping over 6% annually. The overall space is still growing, but the data shows that it’s all due to Hyperconverged infrastructure (HCI).
Then why celebrate? The same reason why we celebrated in the FlexPod team. It’s still an important milestone – it marks the passage of time, the accomplishment of another year of business, it’s certainly about the comradery of those in the business ecosystem, and an appreciation that this offering continues to pay the bills for those workers at Dell/EMC.
But is it a time to celebrate for IT customers?
If brains were dynamite you couldn’t blow your nose
Like the ending of American Graffiti, with the main character staring out the window thinking about what might have been, the folks at Dell must have been thinking longingly about the good ‘ol days as they cut into their VxBlock birthday cake. The Converged market for them is a slow glide, riding out the themes of a bygone era of IT. Their website promotes “turnkey converged”, “power” and “mission critical” like it’s a 60’s muscle car. But we’re now solidly within the jet age of Hyperconverged, software-defined infrastructure and cloud-based computing.
I love it when guys peel out
HCI is the new hot rod in town. It’s overtaking all things converged, integrated and “Stack/Block”. HCI is the fastest growing segment in that infrastructure space, and has surpassed the legacy reference architecture and converged segments. It’s currently on a 57% annual ascent and now accounts for almost half of an over $16B market. The customer has spoken: convenience of the more compact form-factor with both storage and compute combined is preferred over the proclaimed benefits of separate “best in class” components or the ability to scale storage resources separately from compute.
For info on an ideal x86 platform for software-defined HCI look no further than the HPE Apollo 4200. It’s built for demanding, data-centric workloads, and currently hosts Windows-based HCI and vSAN environments. Continue to watch this space for more news on new offerings in the area coming from Microsoft and HPE.
Memory-Driven Flash is a new class of enterprise storage that combines the benefits of memory-speed storage media with the traditional economics and manageability of a hybrid flash architecture.
HPE Memory-Driven Flash is being used to enable business databases and applications to achieve new levels of performance while adding virtually no additional administration. Specific studies are showing Microsoft SQL Server database reads are up to 50% faster, which in turn speeds transactions and improves end-user experience.
This architectural approach uses new Storage Class Memory (SCM) as a caching tier to store frequently used blocks of data and enable access to them at near Memory speed, rather than requiring the system to read the data from the storage media tier. HPE’s productization has involved tiering back-end storage media with a layer of Non-Volatile Memory Express (NVMe) solid state drives (SSD), which adds persistent, near-memory performance on top of all-flash SSD. SCM represents the industry’s latest innovation: a low-latency persistent storage media that bridges the performance gap between DRAM system memory and NAND SSD storage. The SCM price/performance level is 10X beneath DRAM and as much as 100X above NAND.
Cache in on the MDF opportunity
In some ways MDF is the modern day evolution of traditional hybrid or ‘adaptive flash’ tiering. Whereas in the past the tiering was between SSDs and spinning disk hard drives, this new instantiation is SCM over SSD. And just as with the previous generation of systems, the unique value of a manufacturer’s product was both superior mechanical design, but also in the intelligent algorithms that identified the ‘hot’ frequently accessed data to cache and optimally managed the precious cache capacity. The image below illustrates how with MDF, the SCM layer can serve up Reads directly from this higher-performance media, rather than accessing SSD. The higher the SCM hit rate, the lower the latency. This approach primarily serves to reduce latency rather than increase system IOPS or some other measure of performance.
NVMe is an open logical device interface protocol for accessing nonvolatile storage over the PCIe interface. The low latency NVMe protocol bypasses and eliminates the overhead of other standard storage protocols such as SAS or SCSI. The NVMe SCM device used in HPE 3PAR storage systems has the NVMe PCIe controller interface as part of the unit, so the NVMe SCM is attached directly to the PCIe bus, and uses the NVMe protocol for faster communication with the HPE 3PAR controller.
Cold, hard cache
The math behind the MDF caching is to maintain data on a higher-performing tier of Storage Class Memory that can deliver Reads at an average of just 10 microseconds. This saves considerable time, in preventing the need to access a data tier of solid state storage that may deliver Reads at 90 microseconds (and compare that to a hard drive tier that would take 10 thousand microseconds to perform that same Read operation).
The results of MDF can be seen both from the perspective of the array itself or the host system accessing the data. In either case, when the Storage Class Memory feature is engaged and populated with hot data, the benefit of reading data from the cache has been shown to reduce system latency by 50% on average.
The chart below shows the results achieved in the lab with Microsoft SQL Server 2017 running on a ProLiant Gen10 server and an HPE 3PAR 9450 array with Storage Class Memory.
From the host perspective, in this test there was shown to be a 59% decrease in latency (sec/Read) with Memory-Driven Flash compared to when the NVMe SCM is disabled. Decreased latency translates to faster database query responses.
From the admin and user perspective, Memory-Driven Flash is simple to use. It’s an embedded component of the system – another storage device that serves as a selectable level of cache in the storage system.
How to Cache in on MDF
Memory-Driven Flash will provide a substantial reduction in I/O response times for small block OLTP type read intensive workloads. However, it will not be beneficial for write intensive or large sequential read or write workloads. Here are some specific guidelines to keep in mind for MDF:
- Memory-Driven Flash does not requires any special SQL Server configuration
- Memory-Driven Flash works with one or more groups of volumes – for best results only enable volume sets that need the boost in performance
- Memory-Driven Flash benefits random read data sizes of 4 KiB to 64 KiB, such as read intensive OLTP type workloads – SQL Server workloads have pre-fetch I/O sizes that are typically 8 KB
- Memory-Driven Flash does not benefit write I/O response times
- Memory-Driven Flash can benefit multiple applications and virtual volume sets concurrently
Get the details on MDF
Track the latest news and happenings with HPE Microsoft Storage solutions on twitter at @mhardi01.