After this release, you can use Spider immediately by installing MariaDB.
For using Spider after installing MariaDB, please execute "install_spider.sql"
which under share directory.
MariaDB 10.0.4 is still alpha version, so please don't use on production systems.
Currently, MariaDB bundled Spider has the following difference than MySQL bundled Spider.
batched key access
This is a feature for join performance improvement.
You can use BKA by setting join_cache_level to 5 and more.
This is a feature for taking table structure from data node.
For example, if there is a table structure as the following
you can create a Spider table by using the following DDL.
CREATE TABLE `t1` (
`id` bigint NOT NULL,
`c1` varchar(100) NOT NULL,
`c2` datetime NOT NULL,
PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=latin1;
CREATE TABLE `t1` ENGINE=SPIDER DEFAULT CHARSET=latin1 CONNECTION='database "test", table "t1", server "s1"';
Current limited features
Slave retring feature by using slave_transaction_retry_errors
Using Spider table through handlersocket
Optimization for single table updating
Partitioned Spider table's
- engine condition pushsown
- fulltext index search
- spacial index search
- full table count optimization
- parallel search
I and Groonga Developer Team were won the 8th Japan OSS incentive award after releasing Spider 3.0.
Thank you for your cooperation!