Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 18 Next »

GEODI can also evaluate databases, i.e., structured data. Databases and their file information (including embedded) can be indexed (all tables and rows). With simple rules, you can limit tables, rows, or fields and set how rows are displayed to the user.

Use Project Wizard/Database connection to define the connection.

Conditions for connection

  1. Connection parameters (connection string) to the DB for readonly access

  2. Find the DB name on the table below if an additional driver required (like ODAC for Oracle)

  3. Sampled data discovery decision (N record per table)

  4. Decision for Indexing and discovery of Embedded or linked files ya da linkli dosyalar

Use “New Connection” to choose the database and define the connection string.

DBMS Driver Requirement List

Some databases may require an extra driver to be installed. Please check the list below.

 DBMS Driver Requirement List

VTYS

Client

PostGres

No extra installation is required.

Microsoft SQL Server

No extra installation is required.

SqLite

No extra installation is required.

Shape File

No extra installation is required.

CSV File

No extra installation is required.

KML File

No extra installation is required.

SQLCE, SQL express

No extra installation is required.

Kafka

No extra installation is required.

Oracle (OleDB)

Oracle ODAC driver must be installed((12.1.0.2.1 and up).

https://www.oracle.com/database/technologies/odac-downloads.html

Excel

Access Database Engine 2010 must be installed.

https://www.microsoft.com/en-us/download/details.aspx?id=13255

MS Access

Access Database Engine 2010 must be installed.

https://www.microsoft.com/en-us/download/details.aspx?id=13255

DB2

Oledb Driver must be installed.

https://www.microsoft.com/en-us/download/details.aspx?id=100917#:~:text=Microsoft%20OLE%20DB%20Provider%20for%20DB2%20works%20with%20Microsoft%20SQL,tools%20to%20IBM%20Db2%20databases.

Oracle BigData

Microsoft Hive ODBC Driver must be installed.

https://www.microsoft.com/en-us/download/details.aspx?id=40886

Cassandra

OLEDB or ODBC driver is required

https://www.cdata.com/drivers/cassandra/download/

MongoDB

OLEDB or ODBC driver is required

https://www.devart.com/odbc/mongodb/

Alternative Connection Methods

Other than using the project wizard

  1. File-based ones such as SQLite, MDB, and AccDB are indexed automatically when found in a directory and don't have a password. You may define DBMeta for them.

  2. You must use Project Wizard for Excel files to be indexed as structured content.

  3. Settings made with Project Wizard/Database can be saved in *.xDeceConnection format. These files are automatically processed. It is a portable and secure way of defining connections.

Indexing DBs

  1. The default behavior is to crawl all tables and rows.

  2. Tables must have primaryKey by default. You may choose to index tables w/o a primary key also.

  3. You may choose a sampling mode, as only the selected number of rows is indexed in each table.

  4. You may choose a subset of tables.

  5. You may limit columns by their names.

  6. You may define SQLs to change content to be indexed

  7. Embedded files are indexed with proper definitions

  8. File paths are processed with proper definitions

  9. If rows have some permission information you may utilize it with proper definitions.

  10. When you search in GEODI, Each row is content. You may define the name and appearance of a record database or table-wise.

Trouble Shooting

 No result from the DB sources

  1. Connection string may not be correct. You should test before saving the project. You can do it now.

  2. Check the permissions

  3. Index error logs should be checked. You can examine the errors received in the scan with the developer report via GEODI.

  4. DBmetas may not be correct

    1. Jsettings files should be syntax validated, you can use https://jsonformatter.curiousconcept.com/ site for this validation.

    2. Check the correctness of the column, table and GEODI project names and their names on the Table/Column/Project database.

 I have no DB connection but results shows many rows

GEODI, by default, crawls file based databases like SQlite or Access. These files are processed like office or PDF files. You may set to ignore them.

 Embedded or linked files does not shown

  1. You should define DB meta to do that. This page explains ho to create DB metas.

 Indexing is slow

  1. By default GEODI crawls all table and all rows. You may use sampled discovery, or narrow the target.

  2. Your DB might also be serving another application.

  3. GEODI indexing speed options might be at low, increase it.

  4. GEODI Server might be running on a low capacity machine, with insufficient memory or slow disk. Please check the system requirements page.

 DB row version are increasing

Every reindex (scheduled or manually) after the first index may generate version for tables w/o a primary key. This does not affect the discovery results, or reports bu only the GEODI Index size.

GEODI indexing generates such tables in the error logs. Please check error log to see wihih tables are affected.

GEODI, by default, requires a primary key (PK) column to exist. So unless you set otherwise this situation does not occur. You may also choose to put a PK to your tables(a best practice). A PK column for GEODI is a column with a unique value in each row.

If table has PK but you still get the error, then primary key column name should be registered to GEODI by defining a DB meta.

 Permissions are not affective

GEODI, by default, reads all rows with same persmissions. To set a row based permission please use DB meta.

FAQ

 Can I index embedded files in the tables?

Yes. with a few simple DBmeta definitions you may index those files. In this case GEODI will show row and file separately. You will be able to query the file using DB attributes.

 How can I open CRM application through GEODI search?

It is possible. You index the CRM database and with a simple DB meta and set a link that opens the original CRM apps. CRM app should have a simple URL(query string) that can be build using the attributes.

DB Meta Definitions

  • No labels