1.3 NDB Cluster API Overview: Terminology

Abstract

This section provides a glossary of terms which are unique to the NDB and MGM APIs, or that have a specialized meaning when applied in the context of either or both of these APIs.

The terms in the following list are useful to an understanding of NDB Cluster, the NDB API, or have a specialized meaning when used in one of these:

Backup.  A complete copy of all NDB Cluster data, transactions and logs, saved to disk.

Restore.  Return the cluster to a previous state, as stored in a backup.

Checkpoint.  Generally speaking, when data is saved to disk, it is said that a checkpoint has been reached. When working with the NDB storage engine, there are two sorts of checkpoints which work together in order to ensure that a consistent view of the cluster's data is maintained. These two types, local checkpoints and global checkpoints, are described in the next few paragraphs:

Local checkpoint (LCP).  This is a checkpoint that is specific to a single node; however, LCPs take place for all nodes in the cluster more or less concurrently. An LCP involves saving all of a node's data to disk, and so usually occurs every few minutes, depending upon the amount of data stored by the node.

More detailed information about LCPs and their behavior can be found in the MySQL Manual; see in particular Defining NDB Cluster Data Nodes.

Global checkpoint (GCP).  A GCP occurs every few seconds, when transactions for all nodes are synchronized and the REDO log is flushed to disk.

A related term is GCI, which stands for “Global Checkpoint ID”. This marks the point in the REDO log where a GCP took place.

Node.  A component of NDB Cluster. 3 node types are supported:

• A management (MGM) node is an instance of ndb_mgmd, the NDB Cluster management server daemon.

• A data node an instance of ndbd, the NDB Cluster data storage daemon, and stores NDB Cluster data. This may also be an instance of ndbmtd, a multithreaded version of ndbd.

• An API nodeis an application that accesses NDB Cluster data. SQL node refers to a mysqld (MySQL Server) process that is connected to the NDB Cluster as an API node.

For more information about these node types, please refer to Section 1.4.3, “Review of NDB Cluster Concepts”, or to NDB Cluster Programs, in the MySQL Manual.

Node failure.  An NDB Cluster is not solely dependent upon the functioning of any single node making up the cluster, which can continue to run even when one node fails.

Node restart.  The process of restarting an NDB Cluster node which has stopped on its own or been stopped deliberately. This can be done for several different reasons, listed here:

• Restarting a node which has shut down on its own. (This is known as forced shutdown or node failure; the other cases discussed here involve manually shutting down the node and restarting it).

• To update the node's configuration.

• As part of a software or hardware upgrade.

• In order to defragment the node's DataMemory.

Initial node restart.  The process of starting an NDB Cluster node with its file system having been removed. This is sometimes used in the course of software upgrades and in other special circumstances.

System crash (system failure).  This can occur when so many data nodes have failed that the NDB Cluster's state can no longer be guaranteed.

System restart.  The process of restarting an NDB Cluster and reinitializing its state from disk logs and checkpoints. This is required after any shutdown of the cluster, planned or unplanned.

Fragment.  Contains a portion of a database table. In the NDB storage engine, a table is broken up into and stored as a number of subsets, usually referred to as fragments. A fragment is sometimes also called a partition.

Fragment replica.  Under the NDB storage engine, each table fragment has number of fragment replicas in order to provide redundancy.

Transporter.  A protocol providing data transfer across a network. The NDB API supports three different types of transporter connections: TCP/IP (local), TCP/IP (remote), and SHM. TCP/IP is, of course, the familiar network protocol that underlies HTTP, FTP, and so forth, on the Internet. SHM stands for Unix-style shared memory segments.

NDB.  This originally stood for “Network DataBase”. It now refers to the MySQL storage engine (named NDB or NDBCLUSTER) used to enable the NDB Cluster distributed database system.

ACC (Access Manager).  An NDB kernel block that handles hash indexes of primary keys providing speedy access to the records. For more information, see The DBACC Block.

TUP (Tuple Manager).  This NDB kernel block handles storage of tuples (records) and contains the filtering engine used to filter out records and attributes when performing reads or updates. See The DBTUP Block, for more information.

TC (Transaction Coordinator).  Handles coordination of transactions and timeouts in the NDB kernel (see The DBTC Block). Provides interfaces to the NDB API for performing indexes and scan operations.

For more information, see NDB Kernel Blocks, elsewhere in this Guide..

See also NDB Cluster Overview, in the MySQL Manual.