VirtualInstruments

What Makes Up the VirtualWisdom^® Solution?

The VirtualWisdom solution is a combination of software and hardware that optimizes the performance, availability and utilization of your SAN and virtualized infrastructure.  It substantially eliminates the risk of deploying I/O-intensive business-critical applications.

VirtualWisdom Solution Components

	VirtualWisdom SAN Availability Probe SAN Availability Probe is a software probe that collects status from SAN switches via SNMP. The software probes collect metrics for each port, and these metrics are generated and reported to the VirtualWisdom Server by each SAN Availability Probe. Each SAN Availability Probe monitors all the ports of a particular fibre-channel switch and collects the metrics for each port, (such as the number of frames and bytes), as well as the key faults for each port, (such as loss of synchronization, link resets, link failures, packet discards, and CRC errors).
	VirtualWisdom Virtual Server Probe Virtual Server Probe is a software probe that collects status from VMware servers via VMware’s vCenter Server. It collects data on and calculates 100+ different metrics including CPU, utilization and status, memory utilization, disk I/O requests and capacity, network requests, and overall utilization, allowing for end-to-end performance analysis from virtual machines to the LUNs. It enables VMware administrators to get a comprehensive view of both the performance and utilization of the virtual servers, and when combined with other VirtualWisdom probes, the entire I/O subsystem.
	VirtualWisdom SAN Performance Probe SAN Performance Probe analyzes every frame header on a fibre channel SAN at line rates of up to 8 Gbits per second. The SAN Performance Probe detects application performance slowdowns and transmission errors by measuring every SCSI I/O transaction from start to finish, for every server/volume combination (initiator/ target/ LUN). It is the only true real-time monitoring device in the industry and is essential for proactive SAN monitoring so that IT managers can avoid unplanned downtime.
	VirtualWisdom Server VirtualWisdom Server software resides on a Windows server platform, controlling and collecting metrics from Virtual Instruments hardware and software probes. It manages the database of metrics, handles integration with third party products via SNMP traps, and handles alerting. The VirtualWisdom Server co-resides with the Virtual Server Probe and the SAN Availability Probe software software on the same physical server.  Two important parts of the VirtualWisdom Server are the VirtualWisdom Dashboard and the VirtualWisdom Views reporting capability.
	VirtualWisdom Dashboard VirtualWisdom Dashboard software is a real-time summary display that provides a customized overview of your virtual and physical infrastructure performance and transmission faults.  It relies on the metrics collected from the three types of probes.  The easy-to-use dashboard displays the activity between VMs, servers, storage network, and storage arrays, including measurements for percent utilization, reads, writes, events, and alarms, as well as signal and error status on the probe channels.  VirtualWisdom Views software is an extremely flexible, powerful, and easy to use report writer that allows users to access the VirtualWisdom SAN metrics and alerts from their workstations in the form of reports, graphs, and tables, in real-time or by historical trends.
SANInsight® Hardware The VirtualWisdom instrumentation, measurement, and analysis solution is designed to provide an unprecedented level of visibility into complex, heterogeneous virtual infrastructures deployed with Fibre Channel SANs. System and storage administrators can implement all or some of the SANInsight hardware components to achieve the appropriate breadth and depth of monitoring and analysis that meet their needs. To provide complete real-time path monitoring and analysis across your virtual infrastructure and multiple SAN tiers, instrumentation is essential. VirtualWisdom makes use of instrumentation hardware and SANInsight Traffic Access Points (TAPs) devices, to measure and record all transactions crossing the SAN. The VirtualWisdom Server software collects, aggregates and correlates all of the physical-layer and operational data.
	SANInsight TAP Patch Panel System The TAP Patch Panel System (TPPS) provide a passive, fail-safe access point to Fibre-Channel network traffic on a SAN for failure analysis, problem diagnosis, and performance monitoring. TAPs operate “out-of-band” by transparently diverting some of the signal through the TAP to another port, which provides a copy of the Fibre Channel frame headers. The agent-less TAP has no impact on application or SAN performance and is integrated with a Fibre Channel Patch Panel for simple deployment.
	SANInsight Rover The Rover physical layer switch extends the monitoring reach of each SAN Performance Probe to groups of SAN links. SAN Performance Probes and protocol analyzers can be attached to Rover physical layer switches and access any of the connected SAN links. Rovers provide a way to deploy a tiered monitoring architecture to lower the cost of monitoring and extend the investment in VirtualWisdom.
	SANInsight Protocol Analyzer The Protocol Analyzer is the industry standard used by all SAN Fibre-Channel device manufacturers to develop their products and to resolve SAN storage, switch, and HBA device issues. Connected via a TAP, the analyzer can be used independently by connecting it to an identified problematic link, via a SAN Performance Probe, or it can be used in conjunction with the SANInsight Rover.

What VirtualWisdom^® Monitors & Analyzes

VirtualWisdom^® is unique in the depth of data that is provided about your physical and virtual infrastructure.   The ability to monitor and analyze both real-time and historical trends from an integrated, easy to use graphical user interface sets VirtualWisdom apart from all other solutions. 

Every time additional servers, virtual machines, HBAs, directors, switches, storage virtualizers, and storage arrays are added to an IT infrastructure, the probability that application response times will degrade due to SAN-induced latency grows. The more complex the environment, including the increased use of virtualization technologies, the higher the likelihood of performance challenges and outages.  Comprehensive monitoring of the SAN, from the VM to the LUN is critical to providing the visibility needed to proactively find problems before they affect users or to quickly resolve the inevitable problems.  The complete set of VirtualWisdom® roving or dedicated probes is the best way to solve these challenges. Proactively monitoring the SAN by setting alarms and running periodic reports prevents SAN performance issues and eliminates more serious downtime issues.

VirtualWisdom is the only solution that directly measures the I/O performance effect (latency, bandwidth, re-tries, errors, etc) on the infrastructure of capacity additions, configuration changes, component/device firmware updates, or application changes.  With this new information, your team can immediately identify and resolve performance problems, improving overall application response times, new application roll-outs, and end-user customer satisfaction.

Some of the key metrics that VirtualWisdom monitors and analyzes include: 

• READ & WRITE Latencies
• HBA Queue Depths / Pending Exchanges
• SCSI Reservation conflicts
• Loss of Signal
• Loss of Sync
• Aborts
• CRC Errors
• Class 3 Discards
• Correlation of latencies with number of transactions
• "What if" modeling to predict the effect on application response time, of SAN configuration changes
• Read data access times – worst and average
• Write data times – worst and average
• Latency time-of-day trends
• Latency-to-load relationships
• Size and type of transaction
• Array response times to read requests – worst and average
• Performance time-of-day trends
• Performance-to-load comparison
• I/O Read & Write request sizes
• Fibre-Channel Link Groups to identify Fibre-Channel linkage problems
• SCSI Link Pending Exchanges to proactively detect SCSI I/O slowdowns
• End device conversation by ITL (Initiator, Target, and LUN) to proactively detect failing SCSI devices
• Real-time aggregated queue depth (regardless of storage vendor or device) to ensure storage ports are optimized for maximum application performance

VirtualWisdom Deployment Options

Virtual Instruments makes it easy to deploy VirtualWisdom solutions by offering a range of options to suit different physical, budget, timing, and staffing needs.  VirtualWisdom can be deployed as a product, offering continuous infrastructure optimization, or as a pure service, offering many of the same benefits, but offloading the customer IT staff and freeing up the capital budget for other projects.

How We're Different

Today’s most common monitoring solutions, such as SAN Fabric Managers, Storage Resource Managers, Network Monitors and others, do not provide the in-depth, real-time monitoring of the IO path needed to truly identify whether an application performance problem is caused by the SAN.  These are all useful tools, but they simply can’t provide the essential information needed to maximize the performance, utilization, and availability of your business-critical SANs.

VirtualWisdom is the only solution that can non-intrusively measure and optimize application performance. By collecting the crucial link layer data, as well as the higher level data from the switches and servers, and correlating that data across the entire environment, VirtualWisdom is the only platform that can provide real-time, continuous infrastructure performance data, by measuring actual SAN I/O traffic.

How VirtualWisdom complements your existing IT management tools:

• Continuous Real-time Monitoring. VirtualWisdom adds continuous real time monitoring and filtering that calculates statistics based on seeing "all" the fibre channel frames that are traveling through the virtualized infrastructure and fibre channel SAN. Other products use polling or averaging techniques that simply do not see every transaction and frequently miss the causes of business-impacting problems.

• Instant Identification.  VirtualWisdom instantly proves whether or not the SAN is the cause of application slowdowns, often reducing root cause analysis from weeks/months to hours/days. At best, the management product you are probably using today reports IOPS or MB/s, which are almost worthless metrics for SAN troubleshooting. By far, the best measure of performance is the effect of the SAN on application response time for every transaction. Looking at IOPS or MB/s is like looking at an automobile speedometer, and guessing how long it takes to go to the market and back for a loaf of bread.

• "What If" Analysis.  VirtualWisdom enables you to do “what if” analyses using actual production data, substantially improving the capacity planning and optimization of your SAN and VMware virtualized infrastructure.

• Accurate Data Movement Analysis.  VirtualWisdom adds a dedicated, non-intrusive traffic and protocol approach to monitoring applications to ensure accurate knowledge about the data movement and data integrity throughout the SAN. Though legacy tools call themselves “out of band”, they are referring to the data path, and for non-performance-oriented functions like capacity management, they are indeed out of band. They do not affect capacity reporting, but they do affect performance and performance reporting, because by their very nature, they consume CPU cycles on each of the infrastructure components.

• Event Recording. VirtualWisdom adds event recording and real time capture capabilities and does not rely on “averages” or polling, which often occurs in 1 – 15 minute intervals. Polling and averaging invariably miss meaningful transaction events. Legacy tools rely on resources in the environment to monitor themselves and then they query those resources for what that resource “believes” to be the state of the transactions and its health, which changes from one firmware release to the next, and one product to the next.

• Performance Trending. VirtualWisdom adds performance trending of SAN device components to identify hardware degradation and enables you to preemptively replace components before they actually fail, which dramatically reduces potential outages. SAN components usually don’t fail catastrophically; they degrade or exhibit intermittent problems. It is often difficult to pinpoint problems and bottlenecks in a production SAN, let alone correct them. VirtualWisdom can identify not only a device failure when a device stops responding to the metric collection queries from the software, or when the traffic through the device comes to a halt or when the device logs out, but also a deteriorating or a misbehaving device, by continuously monitoring and comparing to the known good baseline.

• In-depth Fibre Channel Network Statistics.  VirtualWisdom adds the ability to gather in-depth Fibre Channel network statistics such as pending exchanges, to tune queue depths for maximum performance. In the days of direct-attached storage, when the I/O performance bottleneck was disk transfer rate, existing storage array tools would be sufficient to determine where to place data for an optimum balance of performance and cost. But today, in the world of shared storage, it’s much too simplistic to look just at external transfer rates of disk drives, which may have up to a 4 times difference in performance. Not having HBA queue depths set properly can have a 10 times impact on response time. Virtual Instruments gives you the complete picture.

• Examine SAN Latency. VirtualWisdom adds the ability to determine if configuration changes are affecting SAN performance by examining SAN latency. IT organizations can significantly lower the risk of consolidation projects, or of moving data to lower cost storage tiers by reporting on the effect of the change to application Read or Write response times.

• Validate & Accelerate Storage Decisions.  VirtualWisdom enables more effective use of storage tiering, especially in environments using storage virtualization techniques. Most tiering products use IOPS, MB/s or access frequency to recommend moving data to another tier. Only by seeing the actual effect of the move on application response time can you eliminate all performance risks in tiering decisions.

• Remote Replication Visibility. VirtualWisdom adds remote replication performance and network health troubleshooting. Provides visibility into remote replication throughput and response requirements, enabling stress testing of remote replication infrastructure, providing insight into remote link throughput and latencies, and identifying any misbehavior in the remote replication infrastructure either in the primary SAN, the remote SAN, or on the WAN link/equipment.

• Hardware Agnostics.  VirtualWisdom uses hardware probes that do not rely on storage APIs or SAN firmware changes, so it includes support for all FC SAN devices, including legacy devices, other vendor devices, or the latest bleeding-edge components. VirtualWisdom is ideally suited for change management verification and validation because it is resource & device independent.

• Professional Services. VirtualWisdom is available as a Professional Services consulting engagement, to help offload IT staff with our specialists, and to help those who are constrained by budget cycles, or who have other uses for CAPEX.

Virtual Instruments is a leading provider of innovative solutions to instrument, measure, analyze, and optimize storage area networks and virtualized infrastructures. VirtualWisdom provides real time, continuous SAN monitoring that enables the gathering of detailed VM or host-to-LUN statistics from high traffic fabric links to increase performance, utilization, and availability. VirtualWisdom complements your existing tools and provides visibility to all Host/VM/LUN conversations allowing the accurate measurement of application transaction times within the virtualized environment, and highlighting SAN-induced latency issues that dramatically impact application performance and service level agreements.