When we ask Enterprise Customers, what are the key expectations that you have from IT vendors such as Dell? They always say about Performance ( high IOPs, low latency, fast throughput), Availability (99.999% is a common availability measurement for enterprise storage, Automated load balancing, Fault-tolerant clusters), Automation and Orchestration, Always-on systems that are designed to stay available during firmware upgrades, capacity and performance scaling and most importantly it should be optimized for various Enterprise Workloads (these are some of the key expectations from Enterprise Customers, there could be many more).
When Dell Technologies acquired EMC, they acquired various best-in-class products for their widespread enterprise portfolio. One of the star Enterprise Product Family which shines among all the products is PowerFlex. PowerFlex is ideal for high-value databases and workloads (PowerFlex delivers mission-critical Six 9’s availability for high-value workloads), agile private cloud deployments, and datacenter consolidation.
The PowerFlex software-defined storage platform showed high performance, resilience, and scalability well beyond the hopes of any traditional architecture. Some of all the unique features that make PowerFlex THE BEST offering to Enterprise customers are mentioned below.
Unmatched Flexibility and Scale
- Extreme flexibility with architectural freedom
- Broad environment support and flexible consumption and scaling options
Enterprise-Class Performance and Resiliency
- Massive performance and sub-millisecond latency with large resource pools and uniform data distribution
- Mission-critical availability with extensive fault-tolerance and fast rebuild
- Uncomplicated IT operations and lifecycle management, automation and orchestration with PowerFlex Manager
- PowerFlex has an extensive ecosystem of container and cloud automation platforms and tools that include Kubernetes, Red Hat OpenShift, Rancher, Docker, Google Anthos, vRealize Operations and VMware Cloud Foundation (VCF) among others
- PowerFlex supports up to 100Gbps network infrastructure. With large pools of resources, uniform data distribution, and no network bottleneck, PowerFlex provides massive IO performance that scales linearly with storage resources and nodes
- PowerFlex delivers mission-critical Six 9’s availability for your high-value workloads with multiple protection groups and fast rebuilds
- PowerFlex is validated and optimized for a wide range of high-value workloads, Simple and Efficient Operations
PowerFlex Rack and PowerFlex appliances are intended for Enterprise customers
- PowerFlex rack (have Fully integrated rack-scale fabric, Lifecycle management and orchestration, White glove deployment experience) is a rack-scale engineered system with integrated networking that provides linear scalability and enterprise-grade availability. PowerFlex rack is engineered, manufactured, managed, supported, and sustained as one system for single end-to-end lifecycle support
- The PowerFlex appliance (have Massive scalability potential, Simplified deployment and management, Flexible networking options) offers a turnkey experience in flexible form factors. Customers benefit from massive scalability potential without having to compromise performance and resilience
PowerFlex is IDEAL for Enterprise Workloads
PowerFlex is validated and optimized for a broad set of enterprise workloads including including Oracle database, Microsoft SQL Server, SAP HANA, SAS, Splunk, Elastic Stack, Epic, and the list goes on…
During a recent Dell Technologies lab testing, a 6 node PowerFlex cluster running Oracle RAC database was able to deliver over 1 million IOPs with less than 1-millisecond latency, while an 8-node cluster was able to deliver over 10 million transactions per minute with average latency below 1 millisecond. Testing was done for Microsoft SQL Server, Elastic Stack, Cassandra as well. Please read the output in this whitepaper.
PowerFlex has proven performance outcomes for workloads
- 10 Million TPM with an average latency of 1 millisecond with 8RU
- 1 Million IOPs with an average latency of 1 millisecond with 6RU
Microsoft SQL Server
- 8 Million TPM with an average latency of 1 millisecond with 8RU
- 1 Billion indexing events in 3 hrs. with 7RU
Cassandra DB with Kubernetes
- 300,000 operations/sec with 8RU
Testing by Cloud Evolutions verified that the PowerFlex platform is a high-performance solution—both resilient and highly scalable. The PowerFlex platform tested demonstrated rock-steady performance, scalability, and data availability for the Kubernetes stateful application, CassandraDB, using the Dell EMC PowerFlex CSI driver. Please read the output in this whitepaper.
The results showed that the PowerFlex family achieved:
- Total data populated within 32 Cassandra pods: 18 Billion records (~4.5 TB).
- Total records read 320 Million records
- Over 302,000 operations per second (ops/sec) for 100% reads
- Over 271,000 ops/sec for 100% writes.
- Over 260,000 ops/sec for mixed workloads of 50 percent reads and 50 percent writes, with a median latency of 1.2 milliseconds at 150 threads.
- Consistent linear scaling of ops/sec as more CassandraDB PODS was added to the cluster, from 8 to 32 PODS, without impacting latency.
Do you know the PowerFlex family is one of the best platforms to host the OpenShift cluster along with the MongoDB database?
The most prevalent use cases of MongoDB were chosen to demonstrate the Performance of the PowerFlex Family.
Use case 1: Workload C defines a workload that simulates the capture of 100% read operations only. This use case reads the data within the Mongo pods that is written earlier.
Result: A cumulative throughput of 231,781 ops/s is achieved at a sub-millisecond latency for 64 threads on each MongoDB sharded pod
Use case 2: Workload B defines a read-mostly workload that simulates the capture of 95% read and 5% write operations mix
Result: A cumulative throughput of 207,093 ops/sec is achieved at a sub-millisecond latency for 48 threads on each MongoDB sharded pod
Use case 3: Workload A defines an update heavy workload that simulates the capture of session stored recent actions 50% read and 50% write operations mix.
Result: A cumulative throughput of 151,685 ops/sec is achieved at a sub-millisecond latency for 32 threads on each MongoDB sharded.
PowerFlex provides scalable throughput and IO performance for applications while minimizing infrastructure footprint and management overhead. With large pools of resources, uniform data distribution, and no network bottleneck, PowerFlex provides massive IO performance that scales linearly with storage resources and nodes. This makes it a perfect platform for IO intensive, performance and latency sensitive workloads. Additionally, PowerFlex allows easy isolation of key workloads with multiple protection groups. This isolation helps ensure predictable and uninterrupted performance of high-value performance hungry workloads while ensuring enterprise resiliency and availability.
To know more about PowerFlex, please go through my LinkedIn Posts mentioned below
The Dell PowerFlex family, powered by Intel Xeon processors, is software-defined storage that combines hardware and software to serve high-availability, data-reliant applications through scalability, flexibility, elasticity, and simplicity with predictability at scale Link
PowerFlex Manager provides valuable savings with both deployment and ongoing tasks—savings which accelerate as the PowerFlex platform scales. The automation provided by PowerFlex Manager substantially reduces the impact of human errors in large and complex environments. Link
Do you know the PowerFlex family is one of the best platforms to host the OpenShift cluster along with the MongoDB database? Link
PowerFlex is validated and optimized for a broad set of enterprise workloads including Oracle database, Microsoft SQL Server, SAP HANA, SAS, Splunk, Elastic Stack, Epic, and the list goes on…Link
Top Reasons to Optimize Next-Gen analytics applications with PowerFlex…Link
The PowerFlex software-defined storage platform demonstrated high performance, resilience, and scalability well beyond the expectations of any traditional architecture Link
To help ensure an organization’s success on the digital transformation journey, IT needs a highly scalable infrastructure and service automation. We showed that Docker containers, Kubernetes, and the CSI plug-in for PowerFlex systems, in combination, enable fast and easy provisioning of the Big Data Cluster services Link