For reliable protection of your data, you should develop and implement a backup schedule. You should also ensure that you have a set of tested recovery instructions.
Typical schedules call for occasional full backups interspersed with several incremental backups. The frequency of each depends on the nature of the data that you are protecting.
If you use internal backups, you can use the scheduling features in SQL Anywhere to automate the task. Once you specify a schedule, the backups are performed automatically by the database server. See Automating Tasks Using Schedules and Events.
The length of time your organization can function without access to the data in your database imposes a maximum recovery time, and you should develop and test a backup and recovery plan that meets this requirement.
You should verify that you have the protection you need against media failure on the database file and on the transaction log file. If you are running in a replication environment, you should consider using a mirrored transaction log. See Protecting your data against media failure.
External factors such as available hardware, the size of database files, recovery medium, disk space, and unexpected errors can affect your recovery time. When planning a backup strategy, you should allow additional recovery time for miscellaneous tasks that must be performed, such as entering recovery commands or retrieving and loading tapes.
Adding more files into the recovery scenario increases the places where recovery can fail. As the backup and recovery strategy develops, you should consider checking your recovery plan. See Implementing a backup and recovery plan.