Chat now with support
Chat with Support

SharePlex 8.6.6 - Reference Guide

About this guide Conventions used in this guide SharePlex commands SharePlex parameters SharePlex utilities Appendix B: SharePlex environment variables

Confgure replication to maintain a central data store

These instructions show you how to set up SharePlex for the purpose of consolidated replication: replicating from multiple source systems to one central target system. This strategy supports business requirements such as the following:

  • Real-time reporting and data analysis
  • The accumulation of big data into a central data store/mart or warehouse

Supported targets

Oracle and Open Target

Capabilities

This replication strategy supports the following:

  • Identical or different source and target names
  • Use of vertically partitioned replication
  • Use of horizontally partitioned replication
  • Use of named export and post queues

Requirements

  • Prepare the system, install SharePlex, and configure database accounts according to the instructions in the SharePlex Installation Guide.

  • Prepare the environment for replication. See Prepare an Oracle environment for replication.
  • No DML or DDL should be performed on the target tables except by SharePlex. Tables on the target system that are outside the replication configuration can have DML and DDL operations without affecting replication.
  • Each source system must replicate a different set of data to the central target. If any source systems replicate the same data to the central target system, it is considered to be active-active replication, and conflict resolution routines must be used to determine which source "wins" when the same row is updated from two different sources at the same time. To establish active-active replication, see Configure replication to maintain multiple peer databases.

  • If sequences are unnecessary on the target system, do not replicate them. It can slow down replication. Even if a sequence is used to generate keys in a source table, the sequence values are part of the key columns when the replicated rows are inserted on the target system. The sequence itself does not have to be replicated.

Deployment options

You have two options for deploying SharePlex to replicate from many source systems to one target system.

In either deployment, if any source system cannot make a direct connection to the target system, you can use cascading replication for that route to enable SharePlex to cascade the data an intermediary system that allows connection to the target. For more information, see Configure replication through an intermediary system.

Note: The SharePlex compare and repair commands cannot be used in a cascading configuration.

Deploy with one instance of SharePlex on the target system

You can use one instance of SharePlex to process all incoming data on the target. For each source system, SharePlex creates an Import process on the central target system when replication starts. That, in turn, creates post queues and Post processes for each source-target replication stream, all controlled by one sp_cop process. You can control each source-target stream separately, but the post queues all share the same SharePlex variable-data directory on the target system.

A deployment with a single variable-data directory has the following potential risks:

  • An event that interrupts processing to and from the disk that contains the variable-data directory will affect all replication streams.
  • Any cleanup utilities that you use will affect all of the replication streams, because the cleanup is performed across the entire variable-data directory.
  • A purge config command that is issued on one source system also deletes the data that is replicated from the other source systems, because the purge affects the entire variable-data directory. The use of named post queues eliminates this risk, but adds complexity to the naming, monitoring and management of the SharePlex objects in the deployment.

To use this deployment

  • Install SharePlex in the normal manner, with one port number and one variable-data directory on each system (sources and target).
  • Make certain that when you install SharePlex, you create a database account for SharePlex for each installation.
  • Important! Use the same port number for SharePlex on all systems.

Deploy with multiple instances of SharePlex on the target system

You can deploy multiple instances of SharePlex on the target, one for each source system. A SharePlex instance is composed of the following elements:

  • A unique sp_cop process
  • A unique variable-data directory
  • A unique port number on which sp_cop runs
  • A unique database account that the processes of that instance use to interact with the database.

By running multiple, distinct instances of SharePlex, you can isolate each source-target replication stream from the others. It enables you to:

  • Avoid contention problems that can occur if multiple processes must compete for access to the same queues in a single variable-data directory.
  • Purge one configuration, or clean up and resynchronize one replication stream, while allowing the others to continue processing data.
  • Place the variable-data directories on separate disks so that problems with one disk do not affect the variable-data directories on the other disks.

To use this deployment

Install on the target system first, if possible. This enables you to establish a port number for each variable-data directory, which you can then refer to when you set up SharePlex on the corresponding source system.

Steps on the target system

Select either of the setup options presented in Run multiple instances of SharePlex. These procedures will guide you through the steps to establish independent instances of SharePlex on the target. If you already installed SharePlex on the target, a variable-data directory, database account, and port number already exist. You can dedicate that SharePlex instance to one of the source systems, and then create additional instances on the target per those instructions.

Steps on the source systems

Install one instance of SharePlex on each source system, as directed in the SharePlex Installation Guide. Match the port numbers of those instances to the port numbers of their associated target variable-data directories. If you already installed SharePlex on the source systems, you can change the port numbers as needed. For more information, see Set the SharePlex port number.

Configuration

Create a configuration file on each source system that replicates the objects from that system to the central target. For more information about creating a configuration file, see Create configuration files.

datasource_specification

   
source_specification target_specification central_host[@db]

where:

  • source_specification is the fully qualified name of a source object (owner.object) or a wildcarded specification.
  • target_specification is the fully qualified name of a target object (owner.object) or a wildcarded specification.
  • central_host is the target system.
  • db is a database specification. The database specification consists of either o. or r. prepended to the Oracle SID, tns_names alias, or database name, as appropriate for the connection type. A database identifier is not required if the target is JMS, Kafka or a file.
Example

This example shows data from datasource oraA on hostA and datasource oraB on hostB replicating to oraC on system hostC.

Data from hostA
Datasource:o.oraA    
hr.* hr.* hostC@o.oraC
fin.* fin* hostC@o.oraC
Data from hostB
Datasource:o.oraA    
cust.* hr.* hostC@o.oraC

mfg.*

mfg.* hostC@o.oraC

Recommended target configuration

Each source system in a consolidated configuration sends a discrete data stream that flows to its own Post process on the target. You can assign a unique identifier of your choosing to each source system, and then configure the Post process to include that identifier in each insert or update that it posts on the target.

By identifying rows in this manner, your environment is prepared to support the SharePlex compare and repair commands (which require a source ID) as well as any other work that may require the selection or identification of rows by their source. The compare and repair processes will use the source ID value to select only the rows that are valid for that source.

To configure each Post to write a source ID

  1. Create or alter the target table to include a column named SHAREPLEX_SOURCE_ID. This is the column that will contain the source ID value.

    Note: You can change this name by running the target command with the set metadata option, before continuing further. See the SharePlex Reference Guide for more information.

  2. Choose a unique ID for each of the source systems. Any single alphanumeric string is permitted.
  3. On the target, run sp_ctrl for each Post process.
  4. For each Post process, issue the target command with the set source option. This command sets the source ID that will be posted by that Post process. The following example shows the command for three Post processes:

    sp_ctrl> target sys4 queue Q1 set source east

    sp_ctrl> target sys4 queue Q2 set source central

    sp_ctrl> target sys4 queue Q3 set source west

Related Documents

The document was helpful.

Select Rating

I easily found the information I needed.

Select Rating