Blogs

Agentless “Software only” SAN monitoring, Part 2

Part 2 in a 4-part series

By Ravi Prakash, Director Product Marketing

Application latency in a Fibre Channel over Ethernet (FCoE) environment

If your virtualized applications running on Cisco UCS servers experience slowness, this could be caused by throughput issues in the Fibre Channel over Ethernet (FCoE) portion of your network between the host and the SAN switch.

Our agentless SAN integration uses SNMP to collect metrics from FCoE capable switches like the Cisco Nexus or MDS at a granularity of 1 minute.  It can collect metrics on FCS errors, carrier sense errors, transmit and receive pause frame.  The ability to combine these FCoE metrics with those gathered by our agentless vCenter integration allows for close correlation and enables effective root-cause analysis.

 fcoe-agentless-san-monitoring

Agentless monitoring of Cisco SAN at line rate

What if your most business critical “cannot fail” applications require infrastructure monitoring at a few seconds of granularity (versus few minutes of granularity using our agentless SAN integration)?  Cisco SAN Telemetry Streaming (STS) and VirtualWisdom is your solution of choice here.

Cisco introduced a SAN telemetry offload ASIC in the Cisco MDS 9700 48-port 32-Gps Fibre channel switching module which when installed in an MDS 9000 series switch causes telemetry information to be streamed to VirtualWisdom at 10 sec granularities.  The statistical summary of observed workload response times from this ASIC is similar in nature to that produced by Virtual Instruments hardware performance probes.  Now you have the benefits of VirtualWisdom dashboards, analytics, investigations at a line-rate software-only solution.

Cisco-SAN-Telemetry-Streaming

In addition to the MDS 9000 switch, the Cisco MDS 9132T switch also support STS and will work with VirtualWisdom to give you agentless line rate application-centric performance monitoring.

Need context?  Read Part 1! and stay tuned for part 3 of this 4- part series where we’ll cover hypervisor monitoring and the relevance of mapping applications to infrastructure

Read part 3.

Don’t forget to follow us on TwitterLinkedIn and Facebook to stay up to date on the latest and greatest in app-centric infrastructure performance monitoring.