Shadow Paging

As a shadow paging method is called in computer science, in order to support the two ACID properties Atomicity and Durability in database systems.

A page ( Page ) called in this context a unit of physical memory ( for example, a hard disk), usually of the order to bytes.

Shadow paging is a copy- on-write method, to avoid in-place updates on sides. Instead, a shadow page is when a page is to be modified allocated. As show on the Shadow Page no references ( other sites), it can be modified as desired without violating consistency constraints etc. If a page permanently (persistent ) is to be, all pages that are based on the original are updated so that instead they refer to the new page. Since the site is only "activated" when it is ready, the process runs atomically.

If the referencing pages need to be updated by means of shadow paging, this process can be repeated recursively, which can be quite costly. A solution that is in the WAFL file system is used to make the pages only insofar as persistent, as is even ( ie, write -behind caching ) is necessary. This increases the performance significantly, as many write operations are avoided Hotspots high up in the referential hierarchy (eg, the superblock of a file system ). The price is high time delay in the commit.

Write ahead logging is a more popular solution that uses in-place updates.

Shadow paging is similar to the old master new master batch, which is used in databases on mainframes. In these systems, the result of each batch was (possibly one days work) written on two separate disks (or other data storage ). One was kept for backup, the other was used as the starting point for the next batch.

Shadow paging is similar to the purely functional data structures in as much to be avoided as in-place updates.