Every Oracle DBA will eventually encounter the log file switch wait events. To summarize, the log writer background process is ready to switch writing from one online redo log group and begin writing to another, but it is prevented from doing so. As a result, all DML-related activity eventually comes to a screeching halt. There are two main ways this can occur, and each has its own related wait event. Figure 8-18 will be used to clarify both situations.

Figure 8-18. Shown are the log writer (LGWR) and archive (ARCH) background processes, three online redo log groups, and a disk containing all the archived redo logs. The log writer background process is currently writing to online redo log group RLOG 100, and the archive background process is reading online redo log group RLOG 300 and writing it to the archived redo log disk.

To complete a database checkpoint, the database writer must write every associated dirty buffer to disk, and then every database file and control must be updated with the latest checkpoint number. Keep in mind that regardless of the archive mode, Oracle promises that committed transaction data will reside in an Oracle database file, either in .dbf files or redo log files.