Veritas Access Appliance 8.2 Troubleshooting Guide
- Introduction
- General troubleshooting procedures
- Monitoring Access Appliance
- Common recovery procedures
- About common recovery procedures
- Restarting servers
- Restarting cluster services
- Bringing services online
- Recovering from a non-graceful shutdown
- Testing the network connectivity
- Troubleshooting with traceroute
- Using the traceroute command
- Collecting the metasave image of a file system
- Replacing an Ethernet interface card (online mode)
- Replacing an Access Appliance node
- Speeding up episodic replication
- Uninstalling a patch release or software upgrade
- Troubleshooting the Access Appliance cloud as a tier feature
- Troubleshooting Access Appliance installation and configuration issues
- Troubleshooting Access Appliance CIFS issues
- Troubleshooting Access Appliance GUI startup issues
- Troubleshooting Veritas Data Deduplication issues
- Index
About synchronizing an episodic replication job
The first time an episodic replication job is run, Access Appliance makes a full copy of the data from the source location to the destination. Subsequent jobs (triggered manually or through a schedule) only copy incremental changes.
In certain rare cases, data is already present at the destination, but the episodic replication cannot make the incremental changes. Examples of this situation include:
When episodic replication has not been run for several days or weeks, and the changes that are tracked by the VxFS file change log have been overwritten (or possibly corrupted). This log is required for replication.
When an episodic replication job is temporarily disabled and started again, the next job run triggers a full copy of the data.
When some changes have been made to the episodic replication definition. For example, an earlier replication consisted of fs1/folder1, but you want to replicate data in fs1/folder2 also. Because fs1/folder2 requires a full copy, fs1/folder1 is copied once again, even though only incremental changes are needed.
When the direction of the episodic replication has to be reversed from destination to source. Even though most data is present at both the destination and the source, anytime you create a new job at the destination, a full copy is triggered automatically for the first replication.
If an administrator accidentally deletes the internal database for episodic replications and no backup is available, creating a new job (even for an existing configuration) triggers a full copy.
In these cases, instead of waiting to initiate a full copy, you can use the replication episodic job sync command to leverage the existing data at the destination and avoid requiring a full copy. The replication episodic job sync command returns the replication job to a well-defined state and incremental replication can be used.
After you sync a job, the job is re-enabled, and you can use the standard job trigger or set the replication frequency to trigger incremental replication.
Note:
Synchronization is only supported on enabled jobs. If you are not able to resume from a failed job, and you want to use the replication episodic job sync command to recover from this state, follow these steps. First, disable the job, then enable the job again. Then, use the replication episodic job sync command to synchronize the job.
Note:
Synchronization cannot be performed on a paused episodic replication job. If synchronization is performed on a paused job that has been aborted or stopped, the last recovery point objective (RPO) for the paused job is not available.