The server tries to find the anchor.log file each time a location is opened. If anchor.log is not there, the server creates a new one and initializes the new stream of log file names to begin with a0000000.log. If anchor.log is present and the location was shut down normally after its last use, the server will continue writing entries to the end of the last log file created.
If anchor.log is present and the location experienced transient system failure (due to a power outage, operating system failure, etc.) after its last use, the server automatically recovers the location to a logically consistent point. It ensures that all data that was committed before the crash is applied to the location tables, and all other data is rolled back. This recovery occurs when the first user accesses the location. The logging system then continues functioning normally, appending new work to the end of the last log file -- just as if the system had been shut down normally.