Skip to content

How MetaMiner Works

The MetaMiner Server pulls metadata from each BusinessObjects environment on a short cycle, stores it in the MetaMiner Database, and pushes the changes to connected clients.

How it works

sequenceDiagram
  participant CMS as BO CMS
  participant MMS as MetaMiner Server
  participant DB as MetaMiner Database
  participant MMC as MetaMiner Client
  MMS->>CMS: Log on with the service account
  loop every pull cycle (default 5 minutes per object type)
    MMS->>CMS: Query objects changed since the last pull
    CMS-->>MMS: Changed objects
    MMS->>DB: Replace the stored rows for those objects
    MMS->>CMS: Compare object IDs to find deletions
    MMS->>DB: Remove deleted objects
    MMS-->>MMC: Push the change set
  end
  MMC->>MMS: On connect, load the current snapshot
  1. Log on. When the server starts, it logs on to every enabled environment with the account configured in the MetaMiner Administrator. It detects the CMS version and loads the matching BusinessObjects SDK libraries, which ship with MetaMiner.
  2. First pull. The first cycle reads every object of every enabled type. On a large CMS this takes longest; later cycles are incremental.
  3. Incremental pulls. Each later cycle asks the CMS only for objects whose last-modified time is newer than the last one stored. Each object type has its own period; the default is 5 minutes.
  4. Deep processing. Web Intelligence documents, universes (UNV and UNX) and Crystal Reports are opened and parsed for their queries, objects, data providers and SQL. This runs in the background after the shallow pull.
  5. Deletions. Each cycle compares the object IDs in the CMS with those in the MetaMiner Database and removes objects that no longer exist. Objects that appear without a newer timestamp (for example content imported with promotion tools) are found by a full comparison that runs every 15 minutes.
  6. Snapshot to clients. The server builds a snapshot of each environment from the database. A client loads the snapshot when it connects and then receives only the changes.

Optional sources

Source What it adds Configured in
Audit database Usage history, user activity, history of deleted content. First pull reads up to 550 days back by default; later pulls every 20 minutes. Audit and CMS Databases, Audit settings
CMS database (read-only) Faster change and deletion detection on large systems by reading the CMS repository tables directly Audit and CMS Databases
RESTful Web Service Web Intelligence document processing on current BI 4.x releases Document Processing
Server metrics BusinessObjects server status and metrics every 300 seconds Pull settings

Key terms

Term Meaning
Pull cycle One pass over the object types that are due
Shallow pull Object properties from the CMS (name, owner, folder, dates, kind)
Deep processing Content parsed from inside documents and universes
Do-not-disturb window A time range in which a given object type is not pulled
Snapshot The in-memory view of an environment the client loads

MetaMiner reads, it does not write, during a pull

Pulls only read the CMS, the audit database and the CMS database. Changes to BusinessObjects content happen only when a user runs an action (for example remap or promotion) from the client.