Restore the Sybase Unwired Platform consolidated database and transaction logs from backup after complete failure.
As discussed in Failure and Recovery Scenarios, if only one of C:, D:, or E: drives fail, recovery should be automatic once you have completed the appropriate tasks.
These steps are required in case of complete failure of all three drives:
- Restore the computer's C:, D:, and E: drives from backup.
- Delete the *.db, *.log files from their normal places after you have restored the file system.
- Copy the *.db and *.log files from Computer B's backup directory to the normal locations on Computer A.
Note: Copy the *.log files twice—once to the normal transaction logs directory, and once to the mirrored transaction logs directory.
- Restart Sybase Unwired Platform.
- If there have been package deployments or other cluster-affecting operations since the last database backups, the file-system data corresponding to the packages may be out-of-sync with the database contents related to these packages. If this has occurred, the Sybase Unwired Platform servers cannot fully start. The ml.log file indicates a mismatch between the local cluster version and that of the current cluster. To recover from this:
- Choose one of the Unwired Servers.
- Shut down that server.
- Edit the sup.properties file (UnwiredServer directory), and change the cluster.version property value to match that of the current cluster as reported in the logs.
- Run
updateProps -r
from the same directory to apply the change into the clusterdb.
- Restart that Unwired Server.
Note: This server should be able to take over as the Sybase Unwired Platform primary. Sybase recommends you redeploy all your packages (using the UPDATE option) to make sure all the packages you expect to be available really are.
All of your Sybase Unwired Platform clients should delete their UltraLite databases, and perform full synchronizations.