Teradata Architecture

Teradata Architecture:-

To know more about various Teradata we would first need to understand the architecture of Teradata. So let’s look at the Teradata architecture:-

Above is the Teradata architecture. It is simple and easy. It’s all about parsing engine, bynet, amp’s and vdisc (virtual disc). Let’s discuss above architecture in details:-

PARSE ENGINE (PE):-

Parse engine is also called as PE. It’s a virtual processor (vproc). It’s a one man army for Teradata we can say. The PE consists of Session control, Parser, Optimizer and dispatching. Most of the work done by PE only.

Session control:-

The major work of session control is to manage LOGON and LOGOFF. While logging in it check’s user id and password is proper and valid and accordingly returns a connection successful or an error message.

Parser:-

The main work of Parser is to check whether the syntax is correct and also checks whether the tables are exist in data dictionary. Next Parser will check the availability of the columns listed in a query. It also checks the privileges to the users to access the objects.

Optimizer:-

                                The main work of optimizer is to develop least expensive plan to return the result.

Dispatcher:-

                                The main work of dispatcher is to return the data to the users with the help of bynet.

BYNET:-

                                 In simple words, bynet is communicator between PE and AMP’s. All the communication                                  between PE and AMP’s done via bynet. 

AMP:-

                                AMP stands for Access Module Processor.  The main job of each AMP is to manage                                     some portion of the data of each table. AMP’s do the physical work like generating                                         result set, sorting, aggregating and formatting.

Read Also:-