Yes. You can set what percentage of your daily API limit the product should use before stopping replication to ensure that you don’t exceed the limit and other API applications are not impacted.
None. We use the API server, not the UI server. Salesforce is a multi-tenant environment, and each API server is shared with thousands of their customers, so the net impact of replication on overall system load and performance is negligible.
Zero footprint. Nothing is installed in Salesforce.
You will need Java 8, Tomcat 7, 8 or 9, and the product WAR file.
Relational Junction is a Java application, and is therefore completely cross-platform. Deployment options include
- On-Premise servers and personal computers, including Windows, Mac, and UNIX
- Amazon AWS
- Microsoft Cloud
- Any Cloud computing environment where you have access to install Java and Tomcat
All of these databases have to set to accept UTF-8 compliant character sets to prevent corruption of non-ASCII characters.
- Actian Vectorwise
- Loaded via the bulk loader.
- DB2 has two severe limitations:
- Tables may contain only 500 columns
- The potential width of all data must not exceed the database block size, which can be no greater than 32k, whether or not even the potential block size can be exceeded. Other DBMS platforms can create tables with the potential to create very verbose data but will fail at record insertion time only if the actual data exceeds the block size. DB2 won’t even let you create the columns that would allow you do to this. Since Salesforce objects typically have so many large text fields, nearly all text fields have to be defined as CLOB data types instead of VARCHAR, which means the record is physically stored in many separate database blocks, increasing the I/O and decreasing performance.
- Loaded by the bulk loader.
- MySQL has two limitations that can usually be dealt with:
- A single row cannot contain more data than the block size, but you make CLOBs out of the large VARCHAR fields using the .
- You’ll have to adjust /etc/my.cnf parameters as follows to accommodate your particular needs, using your best judgment as to what the exact values should be:
[mysqld] default-storage-engine=INNODB max_allowed_packet=32000000 innodb_buffer_pool_size=134217728 open_files_limit=10000 max_connections=500
- Excellent DBMS for flexibility. Spans database block size if data exceeds the width of the block. Requires a high level of skill to administer.
- SQL Server
- Excellent DBMS for flexibility. Spans database block size if data exceeds the width of the block. Very easy to administer.
- A single row cannot contain more data than the block size, unless you make CLOBs out of the large VARCHAR fields.
- 16k limit on long text fields, which is a challenge because Salesforce allows 32k fields
Relational Junction for Salesforce will mirror your Salesforce data to Microsoft SQL Server, Oracle, MySQL, PostGreSQL, DB2, Sybase. It doesn't make any difference where the actual database is housed, but the Relational Junction server must be in the same Local Area Network to get reasonable performance. Sending data over the internet one record at a time doesn't scale very well. If you are going to use a collocated service, ensure that the database and the Relational Junction server are not in different data centers. Some of the services do not guarantee this.
Nearly all of our customers using on-premise deployment request and are provided a 2-week trial in which to kick the tires. The product is installed on a webinar with our support staff assisting, and you will get relevant training on the spot.
There is no trial for the Cloud deployment. However, we do provide a webinar demo showing how it works.
You can be up and running within 30 minutes. For the on-premise deployment, you’ll need
- a computer with 16GB RAM and a 4-core processor,
- either Windows or UNIX,
- a UTF-8 character set database
The cloud deployment option is set up within hours of signing a contract.
The product will send you a daily log showing the records copied per object and whether added, updated, and deleted. There are also reject logs if record-level errors happen. You have access to these logs in the on-premise deployment option.