Menu ▾ ▴

Concept Simple Mode

Maruf Çetin

Concept: Simple Mode (Schemaless Direct Store)

Turkce Dokumantasyon | English Documentation

Category: Core Concepts & Architecture
Subsystem: Schemaless Operations (AmberDB::Base)
Entry Type: Architectural Concept

1. Definition and Overview

Simple Mode is AmberDB's lightweight operational mode for using tables without creating or defining schema files (.table).

When simple => 1 is configured (or when accessing unmapped tables), AmberDB bypasses schema validation, secondary inverted index maintenance (.fld, .src, .fac, .srt), and metadata lookups. Records are stored directly into the master Berkeley DB (DB_File) hash table (.db) as native key-value pairs.

This mode is ideal for:

  • Session stores and temporary tokens.
  • Key-value caching layers and rate-limiting counters.
  • Rapid prototyping and one-off CLI utilities.

  • Schemaless arbitrary JSON payloads.

    :::text
    Simple Mode vs Standard Mode

    Standard Schema Mode:
    insert_id() > Validates Schema > Writes .db > Updates .inx, .fld, .src, .fac, .srt

    Simple Mode (simple => 1):
    insert_id() > Writes directly to .db (Zero Index Overhead, Maximum Ingestion Speed)

2. Querying in Simple Mode

In Simple Mode:

  • Primary key operations (insert_id, read_id, modify_id, delete_id, exist_id) run with full ACID safety at maximum $O(1)$ speed.
  • Filtered lookups (field_fetch) automatically fall back to fast C-level sequential table scans (recs_scan).

3. Practical Code Example

# Initialize AmberDB instance in Simple Mode
my $adb = AmberDB->new(
    cfg  => { simple => 1, no_backup => 1 },
    path => { dbase_dir => "./dbstore" }
);

# Insert key-value records without any pre-existing .table file
$adb->insert_id("sessions", "sess_98234a", 1001, "admin", time(), "192.168.1.50");

# Read back directly
my @session = $adb->read_id("sessions", "sess_98234a");
print "User: $session[1], Role: $session[2]\n";

4. See Also

MongoDB Logo MongoDB