McObject has released a DBMS test benchmark application for the Android-based T-Mobile/HTC G1 smartphone. The “TestIndex” benchmark results purport to show that McObject's Perst object-oriented database management system (OODBMS) is faster than Android's default SQLite relational embedded… database.

(Click for larger view of TestIndex screen)

In what McObject says is “likely the first-ever database benchmark conducted on an off-the-shelf mobile phone,” the TestIndex benchmark is said to reflect “typical data management functions” on the two open-source data management systems. On three of the four TestIndex metrics, Perst performed database tasks significantly faster than SQLite, claims the company.

TestIndex benchmarks results: SQLite vs. Perst [Source: McObject]

On the Search test, Perst was said to have performed over six times faster, and on the Delete test it was said to have outperformed SQLite by more than a factor of five. Perst was more than 16 times faster on database inserts, claims McObject. However, SQLite performed 1.51 times faster on Scan benchmarks, according to the single-vendor study.

SQLite is the default, pre-installed embedded database offered with Android. As a relational SQL database, it functions quite differently than the OODBMS approach of Perst. In January, McObject released version 4.0 of its Linux-compatible Perst for Java and .NET. Version 4.0 added persistence for “any” object in both the open-source Perst 4.0 and Perst Lite 4.0. At the time, the company also released sample Android applications such as ContactsIndex (pictured below).

McObject's Android-ready ContactsIndex Perst app (Click for details)

According to McObject, the performance disparity in insert and delete operations is most likely partially due to the lack of explicit transaction support in SQLite's Android API. Perst's edge in searching for records “probably stems from latency in SQLite caused by using Android's Java interface to access the native C language database,” stated McObject CEO Steve Graves.

Other factors were said to include SQLite's overhead of parsing, optimizing and executing the interpreted SQL. “In contrast, Perst is all Java, and its interface works directly with database objects — no interpretation of an intermediate language is needed,” stated Graves. SQLite's victory in scan operations “probably stems from the test's simple tabular data layout,” Graves stated.

Availability

More information on the TestIndex benchmark comparison between Perst and SQLite may be found here.

Perst 4.0 and Perst Lite 4.0 for Java and .NET can be downloaded now with full source code, and should be available here.

---

 
This article was originally published on LinuxDevices.com and has been donated to the open source community by QuinStreet Inc. Please visit LinuxToday.com for up-to-date news and articles about Linux and open source.