(one) Key Compaction: Study each of the static info from disk, merge it Along with the dynamic facts after which create back to disk as new static facts. This can be the most costly algorithm and is usually made use of right after DDL operation.
OceanBase adopts MVCC to do concurrency Regulate. Should the Procedure consists of solitary partition or several partitions on single server node, it will study the snapshot of that server node. If your Procedure entails partitions on a number of server nodes, it executes distributed snapshot examine.
OceanBase adopts shared-very little process architecture. It merchants replica of every partition on not less than three server nodes in numerous server clusters. Just about every server node has its very own SQL execution motor and storage engine. The storage motor only accesses the nearby data on that node. The SQL motor accesses the global schema and generates the distributed query prepare. Query executors pay a visit to the storage engine of each node to distribute and gather info amid them to execute the query.
one.0), as Every single row Model requires further House for desk metadata. In the event the row Edition Restrict is attained, Incorporate COLUMN and DROP COLUMN functions working with ALGORITHM=Fast are turned down having an mistake information that suggests rebuilding the table using the Duplicate or INPLACE algorithm.
ADD COLUMN is not really an in-position operation for saved columns (accomplished without the need of applying a temporary desk) because the expression has to be evaluated because of the server.
OceanBase supports typical SQL query interface, even though there are actually slight differences in OceanBase implementation.
For each databases occasion, it sets a single server node as active root server to offer root services like checking the overall health of all the participant nodes. The basis company is accountable for load harmony, details consistency, error Restoration, and many others. If this Energetic root server shuts down, OceanBase automatically encourages one particular standby root server to a completely new Lively root server.
In other words, a brand new partitioned table is designed with the new partitioning scheme. The freshly made desk includes any variations used by the ALTER TABLE assertion, and desk facts is copied to the new desk structure.
It truly is built and optimized for OLTP apps on relational Oracle迁移OceanBase structured facts, 数据库国产化迁移 however its shared-very little structure also supports OLAP purposes.
general public DDL copyWithPrompts() Gets a replica of this DDL object with all of the statements prompted as appropriate for the command line tool of your native databases.
OceanBase also supports secondary index. It brings together index keys and using表达式优化 desk Major vital for secondary index.
For every database occasion, it sets one particular server node as Energetic root server to supply root provider like checking the health and fitness of all of the participant nodes. The root company is accountable for load balance, knowledge consistency, mistake Restoration, etcetera. If this Energetic root server shuts down, OceanBase routinely promotes just one standby root server to a new active root server.
OceanBase implements code generator to translate the reasonable execution plan into reentrant Bodily execution prepare. Function accomplished by code generator consists of translating the reasonable operators into physical operators, changing the infix expression into suffix expression, leveraging the syntactic information and facts to generate sensible information, reducing the redundant facts structure, and so forth. OceanBase caches these strategies to avoid re-compiling them in future.
OceanBase implements block cache for Micro Block to accelerate big scan query. It also implements a row cache for rows in block cache to speed up compact get question.