Skip to content

Latest commit

 

History

History
250 lines (161 loc) · 8.73 KB

BestPracticeAuditTests.md

File metadata and controls

250 lines (161 loc) · 8.73 KB
Raw

Best Practice Reports

Report 1: HEADER

Header 1: Title

Header 2: General Information

  • Logs: Report name, Report datetime, Workspace name, File size, Creation datetime, Modified datetime, Last Run datetime
  • Warnings about the limit of this procedure.
  • Note current build number.

Report 2: WORKSPACE

Workspace 1: Workspace Version

  • Test which version of FME the last edits were made in compared to the current FME version.
    • If last edits are an older version, WARN about testing before deploying on the current version.
    • If last edits are a newer version, FAIL. Give an ERROR about using a beta build to construct a workspace.
    • If last edits are the same version, PASS.

Workspace 2: Workspace Properties

  • Test for name and category being set.
    • If one or the other are not set, WARN about this being bad for Best Practice.
    • If both are set, PASS.

Report 3: STYLE

Style 1: Bookmarks

  • Test for no bookmarks.
    • If there are no bookmarks, FAIL. Give an ERROR about this being bad for Best Practice.
  • Calculate the number of transformers per bookmark.
    • If there are too few bookmarks (>10 transformers per bookmark), WARN as being bad for Best Practice
    • If there are too many bookmarks (<5 transformers per bookmark), WARN as being bad for Best Practice
    • If there are an appropriate number of bookmarks, PASS.

Style 2: Annotations

  • Test for no annotations.
    • If there are no annotations, FAIL. Give an ERROR about this being bad for Best Practice.
  • Calculate the number of transformers per annotation.
    • If there are too few annotations (>5 transformers per annotation), WARN as being bad for Best Practice
    • If there are too many annotations (<3 transformers per annotation), WARN as being bad for Best Practice
    • If there are an appropriate number of annotations, PASS.

Report 4: BREAKPOINTS

Breakpoints 1: Breakpoints

  • Check for breakpoints.
    • If there are enabled breakpoints, FAIL. Give an ERROR about putting a workspace into production with these.
    • If there are disabled breakpoints, WARN about putting a workspace into production with these.
    • If there are no breakpoints, PASS.

Report 5: READERS

Readers 0: Zero Readers

  • Check for no Readers at all.
    • If there are no readers, bypass other tests. Give a generic No Readers message.

Readers 1: Build Number

  • Check each reader's "Generate Build" number against the current FME version.
    • If the Generate Build was an older version, WARN suggesting upgrade before deploying on the current version.
    • If there are no older versions, PASS.

Readers 2a: Disabled Readers

  • Check for disabled readers.
    • If there are disabled readers, WARN that these are not useful when putting a workspace into production.
    • If there are no disabled readers, PASS.

Readers 2b: Disabled Feature Types

  • Check for disabled feature types.
    • If there are disabled feature types, WARN that these are not useful when putting a workspace into production.
    • If there are no disabled feature types, PASS.

Readers 3: Dangling Readers

  • Check for dangling readers (those without a feature type).
    • If there are dangling readers, FAIL. Give an ERROR that these are positively bad when putting a workspace into production.
    • If there are no dangling readers, PASS.

Readers 4: Excess Feature Types

  • Check for a reader with >10 feature types.
    • If there are readers with >10 feature types, WARN that this indicates the possible need for a merge filter.
    • If there are no readers with >10 feature types, PASS.

Readers 5: Published Parameters

  • Check for basic reader parameters still published.
    • If there are basic reader parameters still published, WARN that these might not be required.
    • If there are no basic reader parameters still published, PASS.

Readers 6: Unconnected Feature Types

  • Check for reader feature types whose name is not a match in the connections list
    • If there are feature types without a match, WARN that these could affect performance and should be removed
    • If there are no unconnected feature types, PASS.

Report 6: WRITERS

Writers 0: Zero Writers

  • Check for no Writers at all.
    • If there are no writers, bypass other tests. Give a generic No Writers message.

Writers 1: Build Number

  • Check each writer's "Generate Build" number against the current FME version.
    • If the Generate Build was an older version, WARN suggesting upgrade before deploying on the current version.
    • If there are no older versions, PASS.

Writers 2a: Disabled Writers

  • Check for disabled writers.
    • If there are disabled writers, WARN that these are not useful when putting a workspace into production.
    • If there are no disabled writers, PASS.

Writers 2b: Disabled Feature Types

  • Check for disabled feature types.
    • If there are disabled feature types, WARN that these are not useful when putting a workspace into production.
    • If there are no disabled feature types, PASS.

Writers 3: Dangling Writers

  • Check for dangling writer (those without a feature type).
    • If there are dangling writers, WARN that these are not useful when putting a workspace into production.
    • If there are no dangling writers, PASS.

Writers 4: Invalid Feature Type Names

  • Check for writer feature type names with possibly invalid characters
    • If there are characters outside of A-Z,a-z,0-9, WARN that these may be invalid (though FME should handle anyway).
    • If there are no characters outside of A-Z,a-z,0-9, PASS.

Writers 5: Mixed Case Attr Names

  • Check Writer Attr Names for instances of mixed-case names (eg some UPPER, some lower, some MixedCase)
    • If there are mixed-case names, WARN that this is not good practice.
    • If there are no mixed-case names, PASS.

Writers 6: Coordinate Systems

  • Check for a writer setting a coordinate system.
  • Check for a reader setting a coordinate system.
  • Check for a CoordinateSystemSetter transformer.
    • If a writer sets a coordinate system, and no reader sets it, and there is no CoordinateSystemSetter, WARN that FME needs to know what to reproject from
    • If there are no writers with coordinate systems, or there are but a reader/transformer sets it, PASS.

Writers 7: Published Parameters

  • Check for basic writer parameters still published.
    • If there are basic writer parameters still published, WARN that these might not be required.
    • If there are no basic writer parameters still published, PASS.

Report 7: Transformers

Transformers 1: Transformer Histogram

  • List the most-used transformers

Transformers 2: Percentage Check

  • Check percentage of overall count
    • Check if there are 6 or more transformers (else the test is not worth doing)
    • If there are >5 transformers and >=25% are all the same type, WARN that this might not be a good scenario

Transformers 3: Transformer Version

  • Check whether an update is required.
    • If a transformer is not the latest version, WARN

Transformers 4: Inspectors and Loggers

  • Check for Inspector transformers
    • Enabled Inspectors = ERROR (note how bad this is for Server)
    • Disabled Inspectors = WARN
  • Check for Logger transformers
    • Enabled Loggers = ERROR (note how bad this is for Server)
    • Disabled Loggers = WARNING

Transformers 5: FMEFunctionCaller

  • Check for FMEFunctionCaller transformers
    • If one exists, it indicates low-level badness. WARN (but a very strong one)

Transformers 6: Similar Parameters

  • Check for transformer parameters that are the same value
    • Suggest these might be better as a user parameter (possibly private) or Custom Transformer: WARN

Report 8: PERFORMANCE

Performance 0: Zero Transformers

  • Check for no transformers at all.
    • If there are no transformers, bypass other transformer-related tests. Give a generic No Transformers message.

Performance 1: Parallel Processing

  • Check parallel processing parameters for use of Aggressive or Extreme.
    • If there are cases of Aggressive or Extreme, WARN that this may not guarantee good performance.
    • If there are no cases of Aggressive or Extreme, PASS.

Performance 2: Parallel Processing 2

  • Check for multiple uses of Parallel Processing.
    • If there are multiple uses, WARN that this may spawn too many processes.
    • If there are not multiple uses, PASS.

Performance 3: Performance Parameters

  • Check for use of transformers with parameters for performance improvements (eg Clippers first)
  • Tests for Clipper/Displacer/FeatureMerger/NeighborFinder/PointOnAreaOverlayer/SpatialFilter
    • If such transformers exist, but don't use that parameter, WARN that this may impact performance.
    • If such transformers exist, and use that parameter, PASS.

Report 9999: SUMMARY

Summary 1

  • List the number of Errors

Summary 2

  • List the number of Warnings