Browse Prior Art Database

Physical Position Anchoring in Treemaps of Hardware Maps Disclosure Number: IPCOM000224430D
Publication Date: 2012-Dec-18
Document File: 5 page(s) / 162K

Publishing Venue

The Prior Art Database


Disclosed is a cartogram-like method to place tree maps within hardware maps and maintain the physical positions of the hardware.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 5

Physical Position Anchoring in Treemaps of Hardware Maps

Tree maps are powerful visualizations that are becoming increasingly popular. They are highly scalable; the user can scale to set sizes without scrolling. Hardware maps are also becoming very popular in the industry. They are desirable because they represent the actual hardware systems, so are very intuitive and natural to use.

Users need to quickly visualize and understand key information at a high level in relationship to the actual hardware. The Hardware maps are a popular foundation view because they are easy to orient and provide good landmarks. Tree maps are a powerful visualization tool as they make it easy for users to grasp and understand data and spot trends.

The invention places tree maps within the hardware maps, and maintains the key relationships and borders between hardware elements - a cartogram for hardware maps. The user is able to display tree maps within the hardware maps for key attributes, such as (but not limited to):

• Workload

• VSs

• Input/output (I/O) usage

• Problems

The novel components of the invention include:

• Rectangles are Nodes. Correspond the outlines of the compute nodes (could also show sub rectangles, such as per processor) to represent the data values

• Bay Anchoring (preferred embodiment). Although the compute node sizes vary per the values of the key attributes, relative node position is maintained, obvious, and orienting. Relative relationships with nodes above and below are maintained. This can be thought of as a cartogram within the hardmap.

• Direct Settings. Borders can be moved to change settings (potentially novel, but not core to this invention)

The top use cases for hardware tree maps are:

• Problem Determination - view log data for all components, alerts

• Predictive Failure Analysis - view thresholds, alerts

• Health Assessment

- FW freshness

- "Suspiciousness" index - aggregation of all anomalies

• Workload Assessment - utilizations are high, but not fully maxed-out

- Storage

- I/O

- Memory

- Central Processing Unit (CPU)

- VSs Figure 1: VSs per Node


Page 02 of 5

• Size - # of VSs (Virtual Systems)

• Color - red = decreasing VS trend, green = growing VS trend

Figure 2: Alternative - Could extend color coding to include the bay numbers

Figure 3: Option: Use actual images and maintain physical aspect ratios for nodes that are not compute nodes (open bays, or storage nodes)


Page 03 of 5

Figure 4: Changing VSs Per a Node

• Decrease Size - Proportionately Migrate off a # of VSs

• Increase Size - Proportionately Migrate on a # of VSs, or enable for the next VS creations

Figure 5: VSs Per a Node and per its CPUs


Page 04 of 5

Example Algorithm - VSs per Node and per CPU (1 of 2):