Skip to main content

StreamShift Documentation

How the assessment score is calculated

In database migrations scenarios, the assessment score is primarily relevant for the source database since the target database is usually empty and new.

The score is calculated based on multiple factors including the type, object structure, schema diversity, cardinality and data characteristics of the underlying database.

In general, given a database migration scenario for a specific source and target combination, the higher the source database assessment score, the larger is the effort to migrate this database with respect to another source database with a lower assessment score.

For example, In an Oracle to PostgreSQL database migration scenario:

Given two Source Oracle Databases SALES, and FIN with the following assessment scores:

Source Database SALES : Score 50000

Source Database FIN : Score 25000

And a target database Azure PostGreSQL, it would require more effort to migrate database SALES to the target PostgreSQL database, as compared to database FIN using the same migration technique.

As you plan, customize and undertake your migration activities, you can run multiple assessments and see how this score iteratively changes. For example, excluding certain tables or schemas from migration and rerunning the assessment should lower the assessment score, and in turn the migration effort.

Score calculation

There are various factors to be considered when planning for a migration. Specifically these factors affect migration in terms of duration, requiring more migration compute resources and special handling of the migration application.

When migrating a specific table, Streamshift has identified following factors affecting the migration:

  1. Table size

  2. Total number of tables

  3. Total number of columns

  4. Total number of rows in a tables

  5. Presence of primary and unique keys

  6. Table containing special column types.

Rather than looking at the above factors individually and evaluating the total cost involved(in terms of duration, migration resources), Streamshift will assign a specific score to each factor and the total value of all these factors are presented as a score. For example, higher the Assessment score results in higher cost of migration in terms of longer duration, higher migration compute resources.

Also migration complexity can be evaluated when comparing the assessment score for a table against its source and target database type. To compute such migration complexity involving various factors such as data types, primary keys, no of columns, a simple representation of total assessment score involving such factors will help us to compute such complexity rather than evaluating complexity against various criteria (or factors) of a given table.

Streamshift assignment of scores are documented in the following link - Assessment Calculation Matrix.

Examples

1. table containing simple data types

Here are the characteristics of the table:

  1. Table containing three columns - ID int, Name VARCHAR2(50), AGE int

  2. Table containing ID being a primary key column

  3. Total number of total rows - 1,000,000

  4. Total size of the table is 100 MB

From the Assessment Calculation Matrix, let us consider the various factors associated with the table:

Factors

Value

Category

Score

Column count

3

Narrow Table

100

Size

100 MB

Small table

100

Row Count

1,000,000

Ultra long table

400

Primary key Column

1

10

Data types

3

Simple

3

Total

613

2. table containing complex data types

Here are the characteristics of the table:

  1. Table containing columns id int, name VARCHAR2(50), age int , resume BLOB, hiring_date DATE

  2. Table containing ID being a primary key column

  3. Total number of total rows - 1,000,000

  4. Total size of the table is 10 GB

From the Assessment Calculation Matrix, let us consider the various factors associated with the table:

Factors

Value

Category

Score

Column count

5

Narrow Table

100

Size

10 GB

Large table

10000

Row Count

1,000,000

Ultra long table

400

Primary key Column

1

10

Data types

3

Simple

3

Date types

2

Complex (Blob, Date)

13

Total

10526

3. Wide table containing simple and complex data types

Here are the characteristics of the table:

  1. Table containing 50 columns: 30 of simple data types, 10 of date, 10 of XML

  2. Table containing three columns being a primary key column

  3. Total number of total rows - 1,000

  4. Total size of the table is 1000 MB

From the Assessment Calculation Matrix, let us consider the various factors associated with the table:

Factors

Value

Category

Score

Column count

50

Wide Table

200

Size

1 GB

Medium table

1000

Row Count

1000

Long table

200

Primary key Column

3

30

Data types

30

Simple

30

Date types

10

Complex (Blob, Date)

230 (200 for XML, 30 for Date)

Total

1690

Computing Migration Compatibility

Computing Migration Compatibility