orm比较

LINQ Implementation Scorecard:
LINQ Implementation Scorecard

Remarks

  • Color bar: Green cells indicate better results; Red cells indicate worse results (the most "red" value is explicitly defined in Landmark column); Yellow cells indicate average results. Color bar is continuous, so all the intermediate colors can be used.
  • LINQ in SubSonic is not tested - currently it is impossible to do this on our standard tests and model, because it does not support references.

Performance scorecard:

ORM Performance Scorecard

Remarks

  • Color bar: Green cells indicate better results; Red cells indicate worse results (the most "red" value is explicitly defined in Landmark column); Yellow cells indicate average results. Color bar is continuous, so all the intermediate colors can be used.
  • Above results are produced for 1000 item sequence. If you're interested in all the numbers, see complete ORMBattle.NET Test Suite output (for 50, 100, 1K, 5K, 10K and 30K item sequences).
  • Cells with dashed frames indicate results produced either with exception to specialized API usage rule (CUD Multiple tests for BLToolkit and NHibernate), or results that are copied from nearby cells because of absence of API required by this test::
    - Absence of native query \ materialization in LINQ to SQL. Or, better to say, LINQ is native API there. Thus results in these cells are copies of LINQ test results.
    - Absence of compiled LINQ queries in NHibernate and SubSonic. Results there are copied from regular LINQ query tests.

Charts

LINQ implementation score

LINQ queries performance Paging performance

ORM materialization performance ORM fetch performance

Create, Update and Delete performance - single optation Create, Update and Delete performance - multiple optations

原文地址:https://www.cnblogs.com/jason819/p/2600464.html