The design principle of these Monitoring Dashboards is clarity through a limited amount of information and great design. Access it from a web browser from anywhere and share it with everyone on your team.

The idea behind the Flip Dashboards is to use these in a wallboard where they rotate one after the other.. It includes exactly three dashboards which is all you will need to get to the point and see the status of the storage.

There are three Dashboard  give an overall status of the SVC/V700/V5000 infrastructure

  • Cluster Performance Overview
  • Cluster CPU and Node Port Load
  • Cluster Cache and Buffer Credit Wait % (BCW)

How to Import Dashboards files (.json)

Make sure you are on the “Home” dashboard. Click on “Home”

A screen will open -on the right you will see “Import Dashboard”.

Click to upload file

 

How to use each dashboard

Description of how to use each dashboard

Description of Dashboards

Cluster IO and Latency

This dashboard shows front-end and back-end performance and latency

The gauges always show the last hour mean value

Pict.: Example Flip dashboard for front end & backend performance monitoring

Please use Ctrl+ and Ctrl- or zoom settings of your browser to scale dashboards to fit on your screen


Cluster CPU and Node Port Load Dashboard

Shows CPU load in percent for system CPU and compression CPU

  • Threshold warning levels are added based on best practices. Yellow: warning level set at 75 - 89%. Red: critical 90% or >

Bottom row shows load on Node Ports

  • SAN port data rate send and receive: send- yellow line, receive- green line
  • Left gauge: last 1hr mean data rate value
  • Threshold warning level set to best practices. Can be adjusted for different port speeds.
    • yellow: warning above 720MB/s
    • red:critical at 800 MB/s

 

 

Cluster Cache and BCW Dashboard

  • Left Gauge
    monitors maximum load on cache partitions
    should always be below 80%

  • Cache effectiveness graph
    Dark green - incoming data into cache
    Light green - outgoing data from cache
    White Line - cache hit%

  • Right gauge
    Cache hit % last hour
    the higher the better

  • Buffer Credit Wait (BCW) % on node ports - Shows blockage in the SAN
    BCW% should always be below 10% - better below 3% when flash or SSD storage is used
    Threshold set to red color in graph - quickly determine if reaches >10%

  • Transfer Latency between storage and server - Shows Communication between storage and host
    Orange Transfer latency line should always be below 0.5 ms
    High transfer latency shows that the server is highly loaded and cannot answer to storage requests in time



  • No labels