Memory and Journal Settings

In Memory Limit

The maximum number of fragments in an in-memory stand. An in-memory stand contains the latest version of any new or changed fragments. Periodically, in-memory stands are written to disk as a new stand in the forest. Also, if a stand accumulates a number of fragments beyond this limit, it is automatically saved to disk by a background thread.

In Memory List Size

The size, in megabytes, of the in-memory list storage.

In Memory Tree Size

The size, in megabytes, of the in-memory tree storage. The in memory tree size should be at least 1 or 2 megabytes larger than the largest text or small binary document you plan on loading into the database. The largest small binary file size is always constrained by the “large size threshold” database configuration setting.

In Memory Range Index Size

The size, in megabytes, of the in-memory range index storage.

In Memory Reverse Index Size

The size, in megabytes, of the in-memory reverse index storage.

In Memory Triple Index Size

The size, in megabytes, of the in-memory triple index storage.

Large Size Threshold

Determines the size, in kilobytes, beyond which large binary documents are stored in the Large Data Directory instead of directly in a stand. Binaries smaller than or equal to the threshold are considered small binary files and stored in stands. Binaries larger the threshold are considered large binary files and stored in the Large Data Directory.

Locking

Specifies how robust transaction locking should be. When set to strict, locking enforces mutual exclusion on existing documents and on new documents. When set to fast, locking enforces mutual exclusion on existing and new documents. Instead of locking all the forests on new documents, it uses a hash function to select one forest to lock. In general, this is faster than strict. However, for a short period of time after a new forest is added, some of the transactions need to be retried internally. When set to off, locking does not enforce mutual exclusion on existing documents or on new documents; only use this setting if you are sure all documents you are loading are new (a new bulk load, for example), otherwise you might create duplicate URIs in the database.

Journaling

Specifies how robust transaction journaling should be. When set to strict, the journal protects against MarkLogic Server process failures, host operating system kernel failures, and host hardware failures. When set to fast, the journal protects against MarkLogic Server process failures but not against host operating system kernel failures or host hardware failures. When set to off, the journal does not protect against MarkLogic Server process failures, host operating system kernel failures, or host hardware failures.

Journal Size

The size, in megabytes, of each journal file. The system uses journal files for recovery operations if a transaction fails to complete successfully. The default value should be sufficient for most systems; it is calculated at database configuration time based on the size of your system. If you change the other memory settings, however, the journal size should equal the sum of the in memory list size and the in memory tree size. Additionally, you should add space to the journal size if you use range indexes (particularly if you use a lot of range indexes or have extremely large range indexes), as range index data can take up journal space. Also, if your transactions span multiple forests, you may also need to add journal size, as each journal must keep the lock information for all of the documents in the transaction, not just for the documents that reside in the forest in which the journal exists.

When you change the journal size, the next time the system creates a new journal, it will use the new size limit; existing journals will continue to use the old size limit until they are replaced with new ones (for example, when a journal fills up, when a forest is cleared, or when the system is cleanly shutdown and restarted).

Preallocate Journals

This setting has no effect.

Preload Mapped Data

Specifies whether memory mapped data (for example, range indexes and word lexicons) is loaded into memory when a forest is mounted to the database. Preloading the memory mapped data improves query performance, but uses more memory, especially if you have a lot of range indexes and/or lexicons. Also, it will cause a lot of disk I/O at database startup time, slowing the system performance during the time the mapped data is read into memory. If you do not preload the mapped data, it will be paged into memory dynamically when a query requests data that needs it, slowing the query response time.

Range Index Optimize

Specifies how range indexes are to be optimized. When set to facet-time, range indexes are optimized to minimize the amount of CPU time used. When set to memory-size, range indexes are optimized to minimize the amount of memory used.