Optimization Scorecard

This report is designed to allow you to easily prioritize your application portfolio for a myriad of optimization activities.  As an example, if you use some of the complexity criteria you could understand which apps in your portfolio would be easiest to migrate versus hardest in terms of complexity.

Standard Templates

Coming Soon

This area will contain preloaded criteria, tags, and weights to produce scorecards without the need of the user to complete the scorecard criteria selection process.

How the Math is Done

All criteria start out with a base 100 points available to be granted.  The criteria type determines how many points of the base 100 an object receives. The product of this is then multiplied against the weight to determine the final score.

There are two primary types of scoring criteria:

  1. Relative
    1. Purpose: provide a relative scaling score regardless of the assessment size.
    2. Description: we will rank all applications from top to bottom based on the selected criteria.  The rank number is then divided by the total number of rank-able items to produce a percentile score.
    3. Final Calculation Example: Percentile * 100 * Weight = Score
  2. Boolean
    1. Purpose: provide a scoring mechanism for unrankable or scalable criteria.  Boolean means that there is a TRUE/FALSE statement as to whether a given object has the selected criteria.
    2. Description: if a given object (device or stack) has the select criteria (e.g. rscore:rehost) then it will receive 100 points (note: if it is a device level criteria the stack will receive 100 pts for every device that matches the selected criteria).
    3. Final Calculation Example: (Count of Objects w/ selected Criteria) * 100 * Weight = Score

For more details on the criteria see the Available Criteria section below.

Available Criteria

Name
Description
Use Case
Type
Calculation
Predictive Cloud Run CostPrioritize application stacks with lower cloud cost according to selected Cloud Provider.  The source of this pricing can be found on the IaaS Cloud Pricing pageEconomicalRelativePercentile rank all apps by sum of hourly cost
High ConnectivityPrioritize application stacks with high amounts of connectivity.  Connectivity being defined as the number of times we've seen this app connecting to any device.  This is the opposite of Low Connectivity.ComplexityRelativePercentile rank all apps by sum netstat connections
Low ConnectivityPrioritize application stacks with low amounts of connectivity.  Connectivity being defined as the number of times we've seen this app connecting to any device.  This is the opposite of Low Connectivity.ComplexityRelativePercentile rank all apps by sum netstat connections
High Number of UsersPrioritize application stacks that connect to many IPs in a chosen location(s).  A distinct IP is used to correlate to a user.  This is the opposite of Low Number of Users.Business ImpactRelativePercentile rank all apps by count of distinct IPs coming from a specific location(s)
Low Number of UsersPrioritize application stacks that connect to few IPs in a chosen location(s).  A distinct IP is used to correlate to a user.  This is the opposite of Low Number of Users.Business ImpactRelativePercentile rank all apps by count of IPs coming from a specific location(s)
Location PrioritizationPrioritize application stacks that have more devices located in a specified location(s).Data Center Lease Running OutBooleanDevices the chosen location(s) get points
Devices Over ProvisionedPrioritize application stacks that have more devices that are over provisioned (not using their available resources).  This is the opposite of Devices Under Provisioned.EconomicalRelativeThis is calculated by pulling all servers in the environment whose 95th percentile CPU or Memory utilization is less than or equal to 50% of their provisioned resources. The CPU and memory percentiles are then added together to form a single utilization metric. This metric is then ranked against the rest of the over provisioned population.
Devices Under ProvisionedPrioritize application stacks that have more devices that are under provisioned (exhausting their available resources).  This is the opposite of Devices Over Provisioned.

Performance

RelativeThis is calculated by pulling all servers in the environment whose 95th percentile CPU or Memory utilization is greater than 50% of their provisioned resources. The CPU and memory percentiles are then added together to form a single utilization metric. This metric is then ranked against the rest of the under provisioned population.
Low Device CountPrioritize application stacks that have a low number of devices in them.  This is the opposite of High Device Count.

Complexity

RelativePercentile rank all apps by count of servers
High Device CountPrioritize application stacks that have a high number of devices in them.  This is the opposite of Low Device Count.ComplexityRelativePercentile rank all apps by count of servers
Large Storage FootprintPrioritize application stacks that are using a large amount of storage.  This is the opposite of Small Storage Footprint.Economical & ComplexityRelativePercentile rank all apps by sum of utilized storage
Small Storage Footprint Prioritize application stacks that are using a small amount of storage.  This is the opposite of Large Storage Footprint.Speed & ComplexityRelativePercentile rank all apps by sum of utilized storage
Stack TagsPrioritize application stacks that have a specified tag key value pair.CustomBooleanStacks having the selected key value pair (Tag Key : Tag Value) get points
Device TagsPrioritize application stacks that have devices that have a specified tag key value pair.CustomBooleanStacks having devices having the selected key value pair (Tag Key : Tag Value) get points