This series of blog posts are dedicated to the performance section for Unisphere for PowerMax 9.0. We have made significant improvements in this particular area in terms of usability and functionality and I wanted to take some time to explain the new functionality and user flows.
I wanted to begin with the Dashboard section where we have our “Pre-Canned” Dashboards which provides you with the ability to interrogate across a certain range of objects which will allow you to troubleshoot a potential performance issue on an array with specific KPI’s concentrating on the most common areas of focus. In total we have 12 Dashboards including Array, Storage Groups, Hosts, Thin Pools, Storage Containers, Port Groups, FE Directors, BE Directors, RDF Directors, Disk Groups, Disk Technology and External Directors. The great thing about dashboards is it takes the hard work out of troubleshooting for customers in that it presents you with a set of charts that will help you identify a potential problem if there is one.
I wanted to focus on the Array and Storage Group Dashboard here and talk a little more about them in greater detail.
For this dashboard the timeframe is set for the last 4 hours in the upper right hand corner but the time range can be expanded as you wish if you need to see earlier timeframe over the last hours, days or months going back over the last 12 months. You can also change the graph style as you wish. These are average values but there is a toggle button whereby you can select a Maximum value if desired.
Also in the upper right hand corner we have some useful functionality including the following:
- Ability to save this dashboard as a custom dashboard or as a template.
- Schedule array report or schedule array overview report
- Run PDF report now, Run PDF report now with data and Run array overview PDF report now.
- Navigate to the charts or analyse section
There are 3 separate views for this dashboard across the upper left hand side including Array Workload, Array Efficiency and Performance thresholds. These have been chose in order to give you a broad picture of how the array is performing at present in terms of its workload and performance.
In the previous screenshot you were looking at an array workload view looking at Host and BE Reqs/sec, Host MB’s/sec, % Cache WP and Host and BE IO’s/sec. In the array efficiency we focus on Subscribed and Usable capacity, Snapshot Capacity and % Metadata usage in order to give you a view into consumption patterns on the array.
In the next example I have selected the Storage Group Dashboard and selected a SG which has been seeing heavy Read and Write activity over the last 4 hours. This dashboard you can see 4 separate tiles our current View SG Workload, I/O Profile, Performance Thresholds and SG Noisy Neighbour. The SG Workload tile gives you an indication of read and write activity that is going to the SG across 6 different charts.
I wanted to spend some time on the new functionality we have introduced in 9.0 with the Noisy Neighbor Dashboard in order to represent potential FA contention.
In our conversations with customers one of the most common issues they encounter in their environments is the Noisy Neighbour scenario. For a given Storage Group the Dashboard will chart key performance metrics, and also detail the relationship between the storage group and the associated front end director and ports. The Dashboard will also show other SG’s that are sharing the same ports and could be contributing to performance issues.
There are 6 charts here with 3 graphs on the top which focus on Reads and Writes. Along the bottom we see the FE Directors and FE Ports that this SG is mapped down. Also we have a complete list of all SG’s that share these mappings so if there were a “rogue” SG which was consuming a lot of resources and causing a knock on performance problem you could easily see who the offender was on this list.
Finally I just wanted to show you templates and custom dashboards. The template I’m displaying here is called MetaData Usage and in my case it’s something I need to monitor closely so I have created a chart with all of our Meta Data metrics and saved this is as a template so I can easily view it any time.
Also in the custom dashboard I can save specific objects such as SG “Tier 1 Trading Application” as it is one of my mission critical SG’s that I need to keep a close eye on so by saving it as a custom dashboard I can access it easily.
This concludes this blog on the dashboard section, next I will cover the charting section where we have made huge improvements in the functionality and usability of the product.
Noisy Neighbor Video Example