ESG Validation

ESG Lab Validation: Lenovo Storage S3200

Author(s): Brian Garrett

Background  

Organizations today must get used to the fact that most of their information—from e-mail, financial records, and employee data to sales orders, video surveillance files, and shipping/tracking details—has gone digital. This is not only true for back-end processes; on the front-end, databases are used to drive web-based operations while file sharing and collaboration are improving productivity, and mobile apps are enabling businesses to easily connect with customers and stay competitive.

But as the value and ubiquity of digital information rises within the business, IT managers struggle with the cost, complexity, and performance challenges associated with managing exponentially growing volumes of digital data. As shown in Figure 1, managing data growth and data protection were near the top of the list of the most-cited IT priorities in ESG’s 2015 IT Spending Intentions Survey.[1] The increasing use of server virtualization is also a top priority as organizations recognize the opportunity to improve efficiency by consolidating applications on fewer physical servers. ESG research also indicates that storage-intensive big data and data analytics initiatives are on the rise with a goal of accelerating business intelligence and customer insight.

Figure 1. Top Ten Most Important IT Priorities

Lenovo Storage S3200

The Lenovo Storage S3200 is a versatile, feature-rich, high-performing, high-density hybrid storage system offering enterprise-class storage management features and performance at mid-range price points. Along with the entry-level Lenovo Storage S2200 controller, these storage arrays easily integrate into any network environment, supporting demanding workloads and providing a simple, cost-effective transition into flash technology.

The S3200’s software capabilities, along with its flash-optimized, high-density hardware platform, cost-effectively addresses performance issues as well as data growth challenges associated with a wide variety of applications and workloads—from general-purpose workloads including databases, file and print serving, web serving, e-mail and collaboration, and virtual server consolidation to specialized workloads including video surveillance, big data and analytics, and media streaming.

Easily adding storage enclosures (HDD and/or SSD) allows organizations to economically meet the growing capacity needs of the business without capital investment—in essence “future-proofing” the business. For hybrid support, SSDs can be added for performance, along with high-density SAS drives for cost-effective capacity—with a lot of price/performance points in between.

Upgrading is simple and convenient, so as needs change, organizations can easily upgrade from the entry-level S2200 controller to the high-performance S3200 controller—all without disruption. In addition, the S3200’s architecture is designed for ease of serviceability, which means adding new host interfaces and swapping out controllers, power supplies, drives, etc., is simple.

The S3200 is an extremely reliable and serviceable hardware platform that’s designed for 99.999% availability and is fully compliant with the stringent NEBS telco industry standard.[2] A highly available dual controller architecture with hot pluggable drives, host interfaces, power supplies, and fans provides enterprise-class levels of fault tolerance at surprisingly affordable prices.

The S3200 is a dual controller hybrid storage array that supports a hybrid mix of up to 48 SAS hard disk drives (HDDs) and solid state drives (SSDs) in a small 2U footprint (see Figure 2). A flexible choice of host interface options are supported including FC, iSCSI, and SAS. With optional SAS-connected expansion that supports up to 192 drives in seven external expansion enclosures, the S3200 supports up to 1,536 TB of raw capacity. A flexible hybrid mix of 10K or 15K rpm SAS, 7.2K RPM NL SAS HDDs, self-encrypting drives (SEDs), and SAS SSDs are supported. Up to eight ports of 16Gb Fibre Channel, iSCSI over 10Gb Ethernet, or 12Gb SAS per dual controller system are provided for server attachments.

Figure 2. Lenovo Storage S3200

The S3200 is a simple to manage, easy to scale, and always available storage solution that was designed with a goal of delivering industry-leading performance at an affordable price. A table-driven management interface provides simplified storage deployment. Pools of HDDs, SSDs, and a hybrid mix of both are supported. After configuring storage into one or more pools, the S3200 kicks in and works automatically, delivering a wide range of enterprise-class features at affordable prices, including:

  • Intelligent real-time tiering with flash hybrid, which improves performance and reduces costs. Frequently accessed 4MB chunks of application data are moved to a higher performing tier (e.g., SSD) for faster performance, and infrequently accessed chunks are moved to a lower cost tier (e.g., near-line SAS) for reduced costs.
  • Rapid RAID rebuild, which reduces the risk of data loss after a drive failure with fast RAID rebuilds. Instead of rebuilding an entire RAID group, only the capacity that has been allocated to storage volumes is rebuilt to restore data protection more quickly.
  • Thin provisioning, which helps customers purchase only the storage they need today, with the ability to add more as application requirements grow—reducing the cost of storage capacity and optimizing efficiency. Instead of reserving and consuming storage when host volumes are created, thin provisioning consumes capacity as needed by applications.
  • SSD read caching, which enables priority access to hot data, accelerating read performance, and providing a quick, real-time performance boost for business-critical applications with SSD configured as a caching layer for frequently accessed read data.
  • Virtualized snapshots, which provide point-in-time copies of live data for quick and easy recovery and test development. All snapshots support full reads and writes, and are standalone or nested, with no setup or preplanning needed—eliminating the need for restore and performance hits.
  • Storage pooling, which virtualizes storage access to multiple drive types, and improves I/O by up to 2.5x without impacting applications. Automated pooling enables rapid, flexible storage provisioning and simple configurations changes, simplifying the definition and management of tiered pools of storage. Stored data is distributed across all drive groups in the pool (wide striping), which helps improve performance, lower latency, and achieve higher volume capacities. When a new drive group is added to the pool, the system performs automatic rebalancing to utilize all drives in the pool for better performance.

ESG Lab Validation

ESG Lab performed hands-on evaluation and testing of the Lenovo Storage S3200. Testing was focused on the value of real-time tiering; high performance, availability, and scalability; and simplified storage management.

Hybrid Intelligence and Efficiency

ESG Lab testing began with a tour of the S3200 management interface, Lenovo SAN Manager, with a goal of providing a quick introduction to some of the key capabilities built into the S3200. A home screen shot that was captured during ESG Lab testing is shown in Figure 3. Note how thin provisioning has been used to provision more capacity than is physically available (green versus blue bars), and pooling has been used to configure three pools of storage with different performance characteristics (performance, standard, and archive). The arrows toward the bottom show how the S3200 is delivering intelligent real-time tiering. In this example, 4MB chunks of “hot” data are moving from the standard to the performance tier, and “cold” chunks are moving to the archive tier.

Figure 3. S3200 Management Interface Home Screen

Figure 4 illustrates the effect of data tiering during ESG Lab testing. The screenshot shows a home screen status summary shortly after starting an online transaction processing (OLTP) database workload. The counters and bar graphs indicate that nearly all of the performance is being served by ten disk drives in the green standard pool. The screenshot on the right, which was taken a couple of hours later, shows how most of the performance is being served by a pair of solid-state drives in the red performance pool.

Figure 4. Data Tiering in Action

Intelligent Real-time Data Tiering

The S3200 moves frequently accessed 4MB chunks of application data to high-performance storage media for faster performance, and infrequently accessed chunks to lower cost media for reduced costs. Working in real time, the S3200  monitors and moves data between tiers of storage in minutes to hours versus competitive tiering implementations that can take hours to days to move bigger, and more inefficient, chunks of application data.

ESG Lab tested the S3200’s data tiering capabilities with a goal of quantifying the benefits of tiering in real time. ESG Lab tested an industry-standard workload that simulates the activity of a typical OLTP database application in a pool with 60 600GB RAID-10 10K RPM hard drives and eight 400GB RAID-1 solid-state drives.[3] Real-time tiering kicked in and started migrating frequently accessed data to solid-state within minutes after starting the test. Aggregate performance increased approximately 2.5x over the course of the next 14 hours. Tiering up to solid-state not only improved the number of database transactions, but it also improved I/O response times, as shown in Figure 5.

Figure 5. Data Tiering: Delivering Faster Response Times for an OLTP Database Workload

 

Rapid RAID Rebuild  

Rapid RAID rebuild reduces the risk of data loss after a drive failure. Instead of rebuilding an entire RAID group, only the capacity that has been allocated to storage volumes is rebuilt. While the level of acceleration narrows as the storage system fills up, it also reduces the risk of data loss after a multiple drive failure to some level. Rapid RAID rebuild also reduces the duration of potential application performance degradation. Figure 6shows how a rapid RAID rebuild on a partially full RAID group completed 14.6x faster than a full rebuild during ESG Lab testing.

Figure 6. Reducing the Performance and Impact of Rapid RAID Rebuilds

Why This Matters

The complexity and challenges associated with meeting the growing storage requirements of the business can  lead to increased capital equipment costs (buying more storage than needed) and operational expenses (spending too much time managing storage). This is especially true within highly consolidated virtual server environments serving a mix of application workloads.

Like the autonomic nature of breathing, IT managers don’t have to think about the benefits of the intelligent software that’s running inside a Lenovo Storage S3200. Capacity, efficiency, and performance are automatically optimized in real time. Rapid RAID rebuilds automatically reduce the performance impact and risk associated with a RAID rebuild after a drive failure. ESG Lab was particularly impressed with the cost-effective real-time efficiency that can be achieved with tiering and a hybrid mix of solid-state and traditional hard disk drives.

 

Cost Effective Performance and Scalability

Next, ESG Lab focused on one of the key strengths of the Lenovo Storage S3200: cost-effective performance and scalability for throughput-intensive workloads. Large 512KB sequential reads and writes were tested with the industry-standard IOmeter utility.[4] The maximum aggregate throughput with an average response time of 30 milliseconds or less is shown in Figure 7.

Figure 7. Lenovo Storage S3200 Peak Aggregate Throughput

What the Numbers Mean

A cost-effective storage solution with high levels of aggregate throughput is ideally suited for bandwidth-intensive workloads including big data analytics, video editing, video surveillance, and content delivery. Content delivery is used to define a broad category of solutions that deliver digital video, image, and audio files over a delivery medium such as a broadcasting network, intranet, or the Internet. The aggregate throughput of a storage solution is one of the key factors to consider when architecting and sizing a content delivery system.

To put the aggregate throughput of the S3200 into perspective, the maximum aggregate read throughput (6.1 GB/sec) was used to calculate the number of streams that can be supported for a number of well-known content types, and the maximum capacity and aggregate write throughput (4.6 GB/sec) were used to explore the number of video surveillance cameras that can be supported. The results are summarized in Table 1.

Table 1. Number of Objects Served

Note: Internet image downloads: 640x480 JPEG, 4-second download/view; Video surveillance cameras: HD1080p, 18fps, H.264, 30-day retention

  • A bit stream rate of 128 Kbps for MP3 files was used to determine that an S3200 storage system has the bandwidth required to support nearly 400,000 song downloads or more than 6,000 high-definition  television broadcast streams being viewed at the same time.
  • A bit stream rate of 5,000 Kbps for a video surveillance camera recording in 1080p high-definition at 18 frames per second was used to determine that the S3200 has the bandwidth required to support up to 7,537 cameras.
  • While high aggregate throughput is an important consideration for data-intensive ingest workloads like video surveillance, the cost effective ability to store huge amounts of data is also a key factor. In this example, ESG Lab determined that a fully populated 192-drive S3200 has the capacity to retain 30 days of video surveillance footage from 1,169 cameras.

Why This Matters

A slow storage system can result in loss of sales, loss of customer goodwill, loss of productivity, and loss of a competitive edge. Adding more hardware to solve a storage performance problem only increases capital equipment costs.

With a peak aggregate throughput of more than 6 GB/second and hybrid tiering that delivers a real-time performance boost of 150% or more, ESG Lab has confirmed that the S3200 delivers world-class levels of midrange-storage-price performance.

Simple to Deploy and Manage

The S3200 simplifies storage deployment, streamlining routine configuration changes with a simple to use, table-driven management interface.

ESG Lab Testing

ESG Lab tested the ease of configuring the S3200 from scratch. Screenshots of the S3200’s management interface, taken at the beginning and end of testing, are shown in Figure 8. Starting on the left with a blank configuration, it took just five minutes and 26 mouse clicks to configure the system from scratch, and define and map a storage volume to a server. ESG Lab was not only impressed with the simplicity of the management interface, but also noted that the home screen provides an intuitive, top-down overview of the configuration, health, and utilization of the system—from hosts, through host ports and overall capacity, to a great view of back-end storage pool consumption.

Figure 8. Quick and Easy Deployment

Next, ESG Lab tested the ability to define and map groups of volumes using an intuitive table-driven management interface. Configuring multiple volumes was completed simply using the table-driven interface shown in Figure 9. In this case, the Add Row button was used to define four volumes, each with a different capacity.

Figure 9. Table-driven Multi-volume Management

The screen shot in Figure 10shows the table-driven management interface that was used to map hosts to volumes. In this example, a host port was mapped to two volumes in one simple operation. Note how hosts can be managed using initiator World Wide Names, user-defined nicknames, or user-friendly groups of host names (e.g., all the hosts and initiators associated with a cluster). This ability to map user-defined groups of volumes to groups of user-defined host ports with a single mouse click dramatically simplifies the definition and modification of complex storage configurations—especially in clustered virtual server configurations.

Figure 10. Simple Grouping and Mapping

Why This Matters

Data growth is driving up the cost of deploying and managing storage systems. As a result, wizard- and table-driven configuration and management has become a welcome area of focus and innovation in the storage industry in recent years.

ESG Lab has confirmed that the simple to use S3200 management interface, Lenovo SAN Manager, dramatically reduces the time, effort, and potential for error inherent in deploying and managing complex storage environments. Configuration of a system from scratch was completed in only five minutes. Device creation and mapping were simplified with the table-driven interface.

 

Highly Available and Scalable

The S3200 was designed with a goal of cost effective high availability and enterprise-class levels of fault tolerance. The modular, fully redundant dual controller S3200 is highly available, field serviceable, and extremely dense with support for a hybrid mix of up to 192 hard disk and solid-state drives.

ESG Lab Testing

ESG Lab validated the reliability and serviceability of the S3200 enclosure by injecting system faults, monitoring the system for alerts, and non-disruptively replacing failed components with an I/O-intensive simulated application workload running in the background. As shown in Figure 11, the following errors were injected during ESG Lab testing:

  • Unplug a redundant FC host interface cable.
  • Unplug a redundant power supply cable.
  • Remove a field replaceable fan.
  • Remove a field replaceable hard drive.
  • Remove a field replaceable controller.
Figure 11. Error Injection Testing

First, ESG Lab pulled one of the drives to simulate a drive failure. An alert was quickly displayed in the S3200’s management interface, which made it easy to identify the enclosure and individual drive that had failed. This visual indication made locating the drive very simple, helping to reduce the chance of human error, and minimizing the time-to-failure resolution.

ESG Lab then tested the serviceability of other components of the S3200 shelf by simulating the failure of host ports, power supplies, fans, and even a system controller. Each failure was quickly identified and resolved. With just one click, we were able to review the event log entries that identified the alerts and the resolutions of component failures, along with the associated timestamps. Figure 12shows the S3200’s management interface and event log entries while identifying the controller failure. In all cases, the application workload continued running without error, and no entries were found in operating system event logs.

Figure 12. Fine-grained Historical Performance Monitoring

Why This Matters

The ability of IT end-users to access information anytime from anywhere is not simply desired—it is expected. Downtime not only hurts business productivity, but in many cases it can also impact customer satisfaction and revenue. Because of this, it’s no surprise ESG research indicates that one of the top priorities among respondent organizations for IT in 2015 is improving data backup and recovery.[5]

The Lenovo Storage S3200 has passed some of the most rigorous certification tests in the industry, e.g., the NEBS Level 3 compliance Telco industry standard. NEBS Level 3 certifies that the equipment is fully compliant with a wide range of NEBS criteria—from environmental conditions to reliability qualifications. In addition to enhanced reliability, a key financial benefit of owning NEBS-certified equipment comes from lower lifecycle costs due to longer product lifetimes and reduced repair and maintenance.

ESG Lab validated that the fully redundant and field-serviceable NEBS-compliant architecture of the S3200 delivers enterprise-class levels of high availability and serviceability at surprisingly affordable price points.

ESG Lab Validation Highlights

  • The S3200 increased the efficiency and ease of managing a cost-effective hybrid mix of SSD and HDD storage.
  • The S3200 started to increase performance with a hybrid mix of SSD and HDD in a matter of minutes for an OLTP database workload.
  • A rapid rebuild of a 10% full RAID group decreased rebuild time 14.5x.
  • The S3200 delivered world-class levels of performance with more than 6 GB/sec of sustained aggregate throughput that is ideally suited for bandwidth-intensive workloads including big data analytics, video streaming, and video surveillance.
  • The intuitive management interface was used to configure the S3200 from scratch in five minutes and 26 mouse clicks.
  • ESG Lab confirmed that the fully redundant and field-serviceable S3200 was highly available as it rode through a barrage of simulated hardware failures.

Issues to Consider

  • Real-world application workloads tend to have more locality of reference than the workload generators that were used during this ESG Lab validation. As a result, ESG Lab is confident that intelligent real-time tiering can be used to exceed the 2.5x performance improvements presented in this report. Improvements of up to 5x are realistically achievable depending on the cache friendliness of an application and the ratio of solid-state to hard disk capacity.
  • While ESG Lab testing of intelligent real-time tiering focused on the performance advantages that can be achieved in real time with a relatively modest amount of SSD, real-time tiering can also be used to reduce costs as it moves less frequently used data to a more cost-effective tier of storage.
  • While the S3200 is not a unified solution that supports network-attached storage (NAS) and block-based storage area network (SAN) protocols, it is often deployed as part of a unified solution with the file system layer implemented in a cost-effective virtual server or appliance.

The Bigger Truth

Explosive data growth and budget constraints are driving a new wave of innovation across the storage industry. Lenovo is riding the crest of that wave with its new S3200 SAN array featuring intelligent real-time tiering, high availability and scalability, fast performance, and simple management. This versatile, feature-rich, high-performing, high-density hybrid storage system delivers world-class performance at affordable prices—especially when compared with the growing number of mid-range storage systems built with a layer of software running on an industry-standard Intel processor. The S3200’s NEBS-compliant, purpose-built hardware was designed with a goal of ensuring 99.999% availability. Designed for easy serviceability and upgradeability, adding new host interfaces, swapping out controllers, power supplies, drives, etc., is simple and straightforward. An intuitive management interface simplifies deployment and typical storage management tasks.

The S3200’s virtualized storage pools enable rapid and flexible storage provisioning, while intelligent data tiering optimizes storage use with intelligent data placement—increasing overall system performance, reducing costs, and simplifying management. The fine-grained 4MB efficiency and real-time frequency of the S3200 provide best-in-class efficiency and performance compared with competitive tiering systems with larger page sizes (e.g., 1 GB) and sluggish monitoring and migration frequencies (e.g., 24 hours). Rapid RAID rebuild significantly decreases recovery time and minimizes risk factors, with rapid data restoration. Non-disruptive upgrades from the entry-level S2200 controller, the easy online addition of storage enclosures, and the flexibility of a hybrid architecture that supports a wide variety of HDD and/or SSD media without disruption allows organizations to economically meet the growing capacity needs of the business without capital investment—in essence “future-proofing” the business.

ESG Lab testing showed significant improvements in response times for workloads mirroring typical business environments such as e-mail, database queries, and online ordering. The S3200 quickly adapted in real time to meet the needs of rapidly changing workloads. We believe customers will welcome Lenovo’s flexible, business-optimized approach to storage that saves time and money, is highly available, delivers great performance at a compelling price point, and eliminates the hassle of dealing with RAID.

 


[1] Source: ESG Research Report, 2015 IT Spending Intentions Survey, February 2015.

[2] https://en.wikipedia.org/wiki/Network_Equipment-Building_System

[3] Test bed details: Lenovo Storage S3200, 6 x R10 5+5 600GB 10K and 4 x R1 1+1 400GB SSD, 4x8G FC, Intel Xeon E5620 @2.4GHz, 12GB RAM, Microsoft Windows Server 2008 R2, 8K OLTP database workload.

[4] http://www.iometer.org/

[5] Source: ESG Research Report, 2015 IT Spending Intentions Survey, February 2015.

ESG Lab Reports

The goal of ESG Lab reports is to educate IT professionals about data center technology products for companies of all types and sizes. ESG Lab reports are not meant to replace the evaluation process that should be conducted before making purchasing decisions, but rather to provide insight into these emerging technologies. Our objective is to go over some of the more valuable feature/functions of products, show how they can be used to solve real customer problems and identify any areas needing improvement. ESG Lab's expert third-party perspective is based on our own hands-on testing as well as on interviews with customers who use these products in production environments. This ESG Lab report was sponsored by Lenovo.

Topics: Storage IT Infrastructure Networking Data Platforms, Analytics, & AI