BlockSize

Block size (in bytes) of the database file.

Follow the value with K (to specify a value in KB).

The value should be a multiple of 2KB: minimum 2K, maximum 64K.

16K

RO

BtreeJoinLimit

Maximum percentage of the index page size for the permanent storage (B-tree) that can be filled in order for a page to be a target for a join operation.

Possible values range from 0 to 50.

40

RW/
Startup

CacheSize

Size (in bytes) of database cache memory for the server.

Follow the value with K (to specify a value in KB), M (to specify a value in MB), or G (to specify a value in GB). For example: CacheSize=40M.

The minimum value (4096K) is applicable only to smallest deployments, the recommended minimum value is 16M.

The access mode is RW for increasing the cache size, and RO for decreasing the cache size.

Although solidDB is able to run with a small cache size, a larger cache size speeds up the server. The cache size needed depends on the size of the database file, the number of connected users, and the nature of the operations executed against the server.

For more information, see Managing the database cache.

32M

see description

DirectIO

If set to yes, the index file uses Direct I/O.

Direct I/O means that operating system buffer pool is bypassed in file I/O.

This parameter is not effective in Windows environments; in Windows environments, database files always use Direct I/O.

no

RW/
Startup

ExtendIncrement

Number of blocks of disk space that are allocated at one time when solidDB needs to allocate more space for the database file.

If each block is 8 KB, the value of 500 corresponds to 4 MB of disk space.

500

RW/
Startup

FileSpec_[1...N]

Location, name, and maximum size of the index file.

In solidDB, the term index file is used as a synonym for database file.

The parameter accepts the following arguments:

The device number defines the physical drive number. The value is used as a hint for I/O threads, allowing the server to perform database file I/O requests in a parallel manner if you split the file into multiple physical disks.

For example:

The N in the parameter syntax signifies the number of the file (if the database file is divided into multiple files and onto multiple disks).

The path must use the conventions of your operating system. For example, in Windows environments, if the path contains white space characters, the path must be enclosed in double quotation marks. If the server runs on a UNIX operating system, path separators must be slashes instead of backslashes.

To achieve better performance, the database file must be stored to a local drive using local disk names to avoid problems with network I/O.

You can also have multiple files on a single disk if your physical disk is partitioned into multiple logical disks and no single logical disk can accommodate the size of the database file that you expect to create.

See Using the diskless capability with SMA and LLA servers for information about using the Filespec parameter with diskless servers.

(see description)

RO

PreFlushPercent

Percentage of page buffer that is kept clean by the preflush thread.

The preflush operations prepare the cache for the allocation of new blocks. The blocks are written onto the disk from the tail of the cache based on a Least Recently Used (LRU) algorithm. Therefore, when the new cache blocks are needed, they can be taken immediately without writing the old contents onto the disk.

25

RW/
Startup

ReadAhead

Number of prefetched index reads during long sequential searches.

When the I/O manager is handling a long sequential search, it enters a read-ahead operation mode. This mode ensures that the next file blocks of the search in question are read into the cache in advance. This improves the overall performance of sequential searches.

4

RW/
Startup

ReferenceCacheSizeForHash

Size (in bytes) of the maximum expected cache size.

Follow the value with K (to specify a value in KB) or M (to specify a value in MB). For example: ReferenceCacheSizeForHash=48M

If you set a value for the ReferenceCacheSizeForHash parameter, this value (instead of the current cache size) is used for calculating the size of the cache hash table. If you know in advance what the maximum cache size is during the server lifecycle, then, to avoid collisions, you can use this parameter to set a matching value for the hash table size.

For more information about the relationship between the hash table and the cache size, see Managing the database cache.

0 (not used)

RW/
Startup

SynchronizedWrite

On UNIX/Linux platforms, if set to no, asynchronous I/0 is enabled.

Asynchronous I/O provides, in general, more performance but it can cause higher variance of response latencies (lower latency determinism).

yes

RW/
Startup