Short answers to the questions that come up most often, each with a link to the page that has the details. For an error message or a process that has failed, see Access and Product Validation, Capture and Source Databases and Apply and Target Databases.
Repstance and Repstance Advanced Edition both support SQL Server and Oracle as a Source and as a Target. PostgreSQL, MySQL, Snowflake, Redshift and AWS S3 are Targets of Repstance Advanced Edition only. With Repstance an Apply Process for them does not run, and its status says [postgresql] DB is not supported, with the name of the database type in the brackets. The databases of the on-premises edition depend on its license key; contact support. See Supported Database Requirements.
An AWS or Azure instance runs up to 10 Capture Processes and up to 10 Apply Processes. In the on-premises edition the limits come from the license key. When the server starts it loads the processes in the order of their ids, and a process beyond the limit is not loaded: it does not appear in the output of status, and the service log says Failed to register Apply process ID=11. The maximum number (10) of Apply processes has been reached (Capture instead of Apply for a Capture Process). prepare process=apply beyond the limit is refused with Failed to prepare apply process. The maximum number (10) of Apply processes has been reached.
On AWS and Azure the Data Comparison and the parameter asyncredo=1 need an instance with at least 32 GB of RAM. See Requirements and Limits and Oracle Capture Parameters. The trial and the on-premises edition are not limited.
Open https://HostName:3000/ in a web browser; port 3000 has to be open. The user is repstance. The password is the token of the server: the Instance ID on AWS and the VM ID on Azure, unless it was changed. See Web User Interface (Web UI).
| Port | Needed for |
|---|---|
| 22 | SSH access to the server (AWS and Azure instances) |
| 3000 | The Web UI, from the computers that have a browser |
| 8797 | repcli on other computers and the REST API |
The Web UI reaches the server on the same machine, so port 8797 has to be open only for repcli on other computers and for the REST API. See INTERACTING WITH REPSTANCE SERVER.
Change the value of token in /opt/repstance/conf/repstance.conf and restart the services (see How to Stop Repstance Services and How to Start Repstance Services). The password of the Web UI, the token of repcli on other computers and the header X-Token of the REST API are this one value. See Configure repcli for Remote Access.
The server is supplied with a self-signed certificate, so the browser warns and repcli refuses it while verifyhttps is 1. For repcli either set verifyhttps to 0 or install a certificate; for the Web UI replace the certificate files of the Web UI. Both are described on Web User Interface (Web UI) and Configure repcli for Remote Access.
Not for the replication itself. A primary key or a unique index is needed for the Data Comparison and its repair, for loading a table in pieces (the split load of the Initial Load) and for the Apply parameter mergetables. See Requirements and Limits and Initial Load.
For the tables that match ddlinclude and do not match ddlexclude. Without ddlinclude no DDL is captured. How the Apply Process handles the statements is set by the parameter ddlprocessing on the Apply Parameters pages. See Capture Objects Specification.
Yes. See Multi-Target Replication and Bi-Direction Replication.
With the Initial Load: the Capture Process copies the tables that are listed in loadinclude (they must also match dmlinclude) when it is run. See Initial Data Migration and Initial Load.
With the Data Comparison, which also repairs the rows that differ. See DATA COMPARISON.
Yes. The server checks the processes every 5 minutes, and a process whose run has failed is run again, at most restartattempts times (the default is 5, 0 switches it off). Each check at which the process has not moved forward uses one attempt, and the count starts again when the process has moved. When the attempts are used up, one e-mail goes to the notification channels of the process (when it has any), and the process stays RUN (Failed) until you run it. See Status of any Capture or Apply Processes.
Create a notification channel in the Web UI and attach it to the process (the parameter notification). See Monitoring Replication.
The processes are registered but idle. Only a process with autostart=1 runs by itself when the server starts, the default is 0. Run the others with run (see Run any Capture or Apply Processes).
| Path in /opt/repstance | What it holds |
|---|---|
| conf | repstance.conf (token, port and certificate files of the server), and one file for each process (capture.{id} and apply.{id}) with its parameters, including the passwords of the databases |
| trail | The trail files that the Capture Processes write and the Apply Processes read |
| log | The process logs and the incident files (see Logs and Information for Support) |
| captureckpt.{id} | The checkpoint of each Capture Process |
| cert | The certificate and the key of the server (port 8797) |
| cli | repcli for Linux, macOS and Windows |
| ui2 | The Web UI |
The trail directory grows with the changes of the Source Database until the Apply Processes have applied them (see the parameter cleanuptrail). A Capture Process stops with Not enough space left on device when less than 250 MB are free in the file system that holds the trail directory. Repstance does not delete the log files. See Housekeeping.
See Backup Repstance Files, How to Update Repstance and Housekeeping.
See Logs and Information for Support.
Prev page:
TROUBLESHOOTING AND FAQNext page:
Access and Product Validation