The Dell EMC portfolio of PowerEdge rack servers is designed to optimize application performance. The servers support intuitive tools to simplify and automate tasks over the entire server life cycle. Every data center and virtual environment has unique requirements, and PowerEdge technology provides the flexibility that is needed to build a scalable vSphere infrastructure.
This Ready Stack supports the following PowerEdge servers:
Embedded in each PowerEdge server is iDRAC9, which provides secure and remote server access for a multitude of common management functions. iDRAC with Lifecycle Controller operates regardless of the operating system state or the presence of a hypervisor. It also offers a complete set of server management features, including configuration, operating system deployment, firmware updates, health monitoring, and maintenance.
iDRAC9 supports various remote connectivity interfaces and protocols, including IPMI, Redfish, SMASH-CLP, and WS-Man, enabling administrators to securely configure, deploy, manage, monitor, and update the server. The iDRAC9 HTML5 Web UI provides secure connectivity through HTTPS. The iDRAC9 Enterprise license includes Group Manager, which provides a one-to-many console experience. With Group Manager, administrators can view and manage a set of servers rather than visually inspect the servers for faults and manage them manually.
The Ready Stack architecture network design consists of these functional groups:
The key building block of the OOB management network is the Dell EMC S4148T-ON switch, which provides forty-eight 10GBase-T ports and multiple uplink port options. Using two S4148T‑ON switches provides redundancy.
The Dell EMC S5248F-ON switch provides LAN connectivity. This switch provides up to forty-eight 25 GbE ports and six 100 GbE ports. Using two S5248F-ON switches—referred to as top-of-rack (ToR) switches—provides redundancy. Built-in NICs directly connect the ToR switches to the rackmount servers.
For FC storage traffic, Ready Stack uses Dell EMC Connectrix DS-6620B switches. The DS-6620B switch delivers up to 32 Gb/s FC performance in a 1U form factor and, by default, has twenty-four 32 Gb/s ports enabled. More ports can be added, enabling up to 64 ports per chassis, through Ports on Demand upgrade kits. Brocade Web Tools, the embedded UI on the Connectrix DS-6620B switches, facilitates the monitoring and management of single or small fabrics, switches, and ports.
The Dell EMC XtremIO X2 all-flash storage array helps realize the promise of a simple, agile, scalable, fully virtualized data center by providing:
A dual X-Brick XtremIO X2 cluster storage array with redundant active/active storage controllers provides sufficient capacity, bandwidth, and high availability for the databases in this Ready Stack design. Each FC port in each XtremIO X2 storage controller is connected to a separate FC switch for high bandwidth and high availability.
The XtremIO X2 array is ideally suited for mixed workloads and database applications. XtremIO iCDM natively provides full copy data management capabilities and integrates production copies with nonproduction copies for critical use cases such as dev/test and newer use cases such as on-demand analytics. The XtremIO X2 array achieves a high degree of space efficiency by using in-memory metadata XtremIO Virtual Copies. The inline data deduplication and compression features can yield significant reductions in database size.
The XtremIO X2 array is based on the X-Brick block, which is shown in the following figure. A highly available, active/active storage component, the X-Brick block contains two independent fault-tolerant storage controllers (nodes) in a compact disk array enclosure (DAE) with 72 SSDs.
Figure 2. XtremIO X2 X-Brick block
An X-Brick block has no single point of failure, as illustrated in the following figure. Each X-Brick enclosure has two SAS DAE controllers with dual connections to each storage controller. Each XtremIO X2 cluster has two NVRAM components to help archive unwritten data to permanent storage in the event of a power failure. Each XtremIO X2 storage controller and DAE is equipped with dual power supplies that are each powered from two separate electrical power circuits, in accordance with XtremIO X2 installation best practices.
Figure 3. XtremIO X2 component redundancy
X-Brick blocks can be clustered to create a large scale-out system that grows linearly in IOPS, bandwidth performance, and capacity as more X-Bricks are added. Each X-Brick block can also scale-up capacity with the addition of SSDs.
An XtremIO X2 cluster can have from one to four X-Brick blocks. A cluster of more than one X-Brick block has dual InfiniBand switches for redundancy, and each storage controller is connected to both InfiniBand switches. The InfiniBand switches are also connected to each other to provide increased bandwidth and redundancy.
The XtremIO Management Server (XMS) WebUI provides advanced reporting and analytics, and automated workflow suggestions, reducing administrator time spent on provisioning and performance tuning.
XtremIO X2 storage is designed and optimized for databases with features that enable DBAs to focus on databases and not worry about managing the underlying storage. Database management systems, such Oracle DBMS, SAP, and Microsoft SQL Server, can operate at peak performance on the XtremIO X2 array. The array delivers high performance for diverse workloads such as online transaction processing (OLTP), data warehousing, and hybrid workloads. It also provides outstanding support for applications such as enterprise resource planning (ERP) and customer relationship management (CRM) systems by delivering high IOPS, network throughput, and transactions per second.
Using XtremIO X2 storage for high-performance database applications provides the following benefits:
In addition, protecting the database is easy with the XtremIO X2 array. In contrast with RAID, XDP requires no design for RAID type, data file capacity, load balancing, and tuning, and XVC provides for easy protection and recovery from any operational or logical corruption. XVC enables the creation of frequent point-in-time copies, according to recovery point objective (RPO) intervals of seconds, minutes, or hours, for use in recovery. Recovery using XVC is instantaneous and does not affect system performance.
The Data Domain system is a disk-based inline deduplication appliance and gateway that provides data protection and disaster recovery in the enterprise environment. It provides the following benefits:
In this Ready Stack certified reference system, the XtremIO X2 array provides significant data reduction though inline deduplication, compression, and iCDM. The Data Domain 6300 complements and enhances these gains by providing storage-integrated data protection, through its own rich set of data reduction features, to support backup, archive, and data recovery.
VMware vSphere is a complete and robust virtualization platform that uses dynamic resource pools for flexible and reliable virtualization of business-critical applications. It transforms a computer's physical resources by virtualizing the CPU, RAM, hard disk, and network controller. It also virtualizes graphics processing units (GPUs) for analytics workloads. This transformation creates fully functional VMs that run isolated and encapsulated operating systems and applications.
In this Ready Stack certified reference system, each of the two OLTP databases is running on its own VM, each of which is hosted on ESXi 6.7U1 on a PowerEdge R940 server. Because deploying high-performance databases is possible without the presence of a virtualization environment, running these databases on VMs is not a requirement. For information about bare-metal versus virtualized infrastructure options in this Ready Stack, see Bare-metal versus virtualized infrastructure.
Red Hat Enterprise Linux offers automation capabilities that are designed to limit IT complexity while enhancing workload security and performance for traditional and cloud-native applications. Although this document focuses on Red Hat Enterprise Linux 7.4, you can use other releases of the Red Hat Enterprise Linux operating system if required.
In this Ready Stack, two production database servers are installed on the operating system. XVC Dev and XVC online analytical processing (OLAP) are XVC development and OLAP databases, respectively, that have been repurposed from the respective production databases. The two-server database forms a cluster. Each instance of Red Hat Enterprise Linux is installed as a bare-metal operating system on a PowerEdge R940 server, as shown in the following figure:
Figure 4. Database server resources and roles
This Ready Stack design also includes two XtremIO Virtual Copies of the databases, both of which are installed on a bare-metal instance of Red Hat Enterprise Linux 7.4 that is installed on a PowerEdge R740 server. The XtremIO X2 array uses the XVC databases for database backup snapshots. Although the design shows the server as local to the production servers, you can also use XtremIO X2 storage features to back up the snapshots to a remote location, as needed.