Is there an unsecured wireless network near you?
Storage and network prodder Virtual Instruments climbs up stack to apps
Wednesday, 06 December 2017 02:25

Virtual Instruments has said it is getting more application-centric with its v5.4 release of VirtualWisdom, its infrastructure performance monitoring and analytics platform, which builds on its Xangati acquisition.

The theory is sound, with the basic real-time monitoring of IT infrastructure components being connected to the performance of high-level applications seen through dashboards that line of business executives and application owners can use.

Virtual Instruments (VI) began life with infrastructure element performance monitoring and fault-finding, looking at things like storage networking cabling and switches. It has extended its performance and optimisation capabilities up the application stack into infrastructure and now says it's extending it upwards again.

This release, it claimed, surfaces performance and optimisation issues to upper-level executives in application terms that matter to them, with connectivity downwards to infrastructure and infrastructure component teams. VI said it provides infrastructure visualisation and management in the context of the application.

Existing monitoring tools are silo-centric, and only provide limited visibility into the various underlying infrastructure components, it argued, claiming that with the reality of today’s highly virtualised, multi-vendor data centres, with thousands of application-level and infrastructure-level services, we need application-aware performance monitoring and optimisation assistance.

It’s a seductive argument.

VirtualWisdom 5.4 collects machine and wire data across physical and virtual hosts, Ethernet and Fibre Channel fabrics, software-defined, converged and hyper-converged infrastructure as well as SAN and NAS storage.

VI says it operates out-of-band and in real-time.

Its 4-way feature set includes:

  • Application service assurance to align infrastructure performance with applications needs;
    • Executive and LOB visibility through exec and application-level dashboards,
    • Tiered service level policies to assure the performance of business-critical applications running on shared infrastructure,
    • Discovering and mapping application usage of dynamic and virtualised infrastructure.
  • Workload and capacity optimisation analytics to proactively manage workloads and capacity from VM to storage array;
    • Optimising end-to-end workload placement across VM, network and storage,
    • Proactively detection of potential performance issues and optimisation opportunities through seasonal behaviour analytics.
  • Problem resolution and avoidance analytics so IT teams can proactively collaborate, troubleshoot and diagnose performance issues;
    • Investigation Runbooks to provide Suggestive Analytics to identify and resolve issues for every alarm type,
    • Chat-ops for cross-team collaboration,
    • Detecting anomalies with automatic comparison to performance baselines and correlating potential root causes.
  • Scalable instrumentation; wire and machine data collection in real-time;
    • Adding visibility of software-defined data centres and hyper-converged infrastructure including VxRail, Nutanix, SimpliVity, vSAN, ScaleIO and Netflow,
    • Expanding wire data support for NAS (SMB) and SAN (FCoE).

The VI technology offers continuous real-time monitoring of application workload across the end-to-end infrastructure. VI stresses that the v5.4 release uses machine learning-based analytics to help users measure and guarantee performance-based service level agreements (SLAs).

In contrast to application stack system monitoring and management tools from infrastructure system vendors, such as Dell and HPE, VI would say VirtualWisdom is vendor-agnostic. This could well appeal to data centre operators with heterogeneous infrastructure component suppliers.


VirtualWisdom structure; click for embiggenmemt

It also claims that its v5.4 software facilitates meaningful discussion and co-operation between application owners, line-of-business execs and managers and infrastructure teams about assuring application performance, controlling costs and reducing risk.

This is all well and good but it is a bold set of claims and the details will need checking against your infrastructure details, looking at, for example, Nutanix's Acropolis and other KVM hypervisor support bods.

We note there is no mention of specific core application support, such as databases and their tuning. VirtualWisdom is an infrastructure-based tool still, and not an upper-level application tuner.

This is the first app-level version of VirtualWisdom though, and VI no doubt has a development roadmap for it. Possibly partnerships with upper-level application providers form part of it.

Nor is containerisation mentioned. That’s not surprising. Tools like VirtualWisdom can only be developed when there’s a stable basic infrastructure set into which to develop and insert performance monitoring sensors, and containerisation is still bleeding edge. Again we might expect partnerships to develop here.

Another development direction might be to look at hybrid cloud support and bring in some form of public cloud performance monitoring and optimisation.

VirtualWisdom 5.4 is available at the end of December, and you can get more information here. The App-centric Analytics module is delivered on a subscription basis by the way. ®

Source: http://bit.ly/2AVJjjO