Examples of MemoryBackend

This class defines a very simple backend that stores its information in memory. This is primarily intended for testing purposes with small data sets, as it does not have any indexing mechanism such as would be required to achieve high performance with large data sets. It is also heavily synchronized for simplicity at the expense of performance, rather than providing a more fine-grained locking mechanism.

Entries stored in this backend are held in a LinkedHashMap<DN,Entry> object, which ensures that the order in which you iterate over the entries is the same as the order in which they were inserted. By combining this with the constraint that no entry can be added before its parent, you can ensure that iterating through the entries will always process the parent entries before their children, which is important for both search result processing and LDIF exports.

As mentioned above, no data indexing is performed, so all non-baseObject searches require iteration through the entire data set. If this is to become a more general-purpose backend, then additional HashMap<ByteString,Set<DN>> objects could be used to provide that capability.

There is actually one index that does get maintained within this backend, which is a mapping between the DN of an entry and the DNs of any immediate children of that entry. This is needed to efficiently determine whether an entry has any children (which must not be the case for delete operations).

Examples of MemoryBackend

Examples of org.nasutekds.server.backends.MemoryBackend