ingresSync Command - synchronize changes between copies of a database
IngresSync can be used to synchronize changes between two or more copies of a database in different installations. It uses the incremental rollforward feature of rollforwarddb to perform the synchronization and manages the copying of checkpoint and journal files from installation to installation as needed. See IngresSync in the System Administrator's Guide for more details on How to Set up IngresSync.
The ingressync command has the following format:
ingresSync --source=<hostname> [ --target=<hostname> ] | --hosts=<hostname>,<hostname>
--database=<dbname> --installid=<xx>
[ --status | --ckpsync --restart | --ckpsync_latest --restart
[--no_incremental] [--strict] |
--jnlsync [--rfp_delay] [--strict] |
--jnlsync_only [--rfp_delay] [--strict] |
--jnlrfp | --incremental_done ]
[--no_compress] [--no_exec] [--ssh=<port>]
--source
Identifies the source host for checkpoint and journal sync operations.
--target
Identifies the target hosts for checkpoint and journal sync operations. For journal sync, each target host database must have the status flag INCR_RFP (incremental rollforward) enabled. Database status can be displayed by using the Ingres infodb utility or the –status flag.
--hosts
Comma separated list of sync hosts, the order is not significant. For checkpoint sync (-- ckpsync) source and target hosts must be identified explicitly with the --source and -- target flags. For journal sync (--jnlsync) hosts may be specified with --hosts because copy direction is determined by database status, the databases with the INCR_RFP status are the targets. If sync direction is changed, the journal sync job need not be altered if it is making use of –hosts.
--database
Defines the sync database name. The database must have the same name and owner in both installations.
--installID
Atwo-character string where the first character must be an uppercase letter and the second character must be an uppercase letter or a number from 0 to 9.
--status
Displays the status of both sync installations and checks if the list of users in the source installation exists in the target. Verifies if the location names and physical paths in the source and target installation match with each other.
--ckpsync
Performs the checkpoint sync operation by copying the checkpoint, database configuration, and dump files from the source to target. The steps to do this are:
• Verify users and locations match.
• Shut down target instance, this is required to replace the target database configuration file with the source.
• Copy checkpoint, configuration and dump files to target.
• Restart the target instance.
• Roll forward checkpoint and enable incremental rollforward on the target.
--jnlsync
Copies latest journals from source to target and performs incremental rollforward by synchronizing the source and target databases.
The process to copy the journals from source to target includes the following steps:
Close the current journal and create the new journal on the Source. The current journal file is opened for writing and therefore not processed during the incremental rollforward processing. Closing the current journal before the sync allows the latest committed transactions to sync on the target.
Copy closed journal files from source to target.
Perform incremental rollforward on the target.
If a database checkpoint has been executed on the source since the last jnlsync, the --jnlsync will detect the new checkpoint and perform an implicit --ckpsync in the next journal sync operation.
Note: This will force a shutdown of the target instance. jnlsync operation without ckpsync does not require the target instance to be shut down.
--incremental_done
Returns the target instance to service. This operation is used to return the target to regular service if the source becomes unavailable (i.e. during a fail over operation) or if a change in the sync direction is required. The normal operating state for the target instance is INCR_RFP.
In this state, the target database can perform incremental rollforward operations and the database is in query only mode and updates are not allowed. Once the --incremental_done step is completed, the database returns to normal service mode, updates are allowed and no additional incremental rollforward steps can be performed.
Last modified date: 08/14/2024