backend update.

This commit is contained in:
Bruce Momjian 1998-06-28 16:35:41 +00:00
parent 8986e609ab
commit b15e7df8c9
1 changed files with 44 additions and 47 deletions

View File

@ -8,30 +8,30 @@ PostgreSQL Backend Flowchart
</H1>
<H2 ALIGN=CENTER>
by Bruce Momjian
</H2 ALIGN=CENTER>
</H2>
<P>
Queries come into the backend via data packets coming in through TCP/IP
and Unix Domain sockets. They are loaded into a string, and passed to
A query come into the backend via data packets coming in through TCP/IP
and Unix Domain sockets. It is loaded into a string, and passed to
the
<A HREF="../../backend/parser">parser,</A> where the lexical scanner,
<A HREF="../../backend/parser/scan.l">scan.l,</A>
breaks the query up into tokens(words). The parser
uses
<A HREF="../../backend/parser/gram.y">gram.y</A> and the tokens to
identify the query type, and load the proper query-type-specific
identify the query type, and load the proper query-specific
structure, like
<A HREF="../../include/nodes/parsenodes.h">CreateStmt or SelectStmt.</A>
<P>
The query is then identified as a <I>Utility</I> function or a more
complex query. <I>Utility</I> queries are processed by a
query-type-specific function in <A HREF="../../backend/commands">
commands.</A> Complex queries, like <I>SELECT, UPDATE,</I> and
<I>DELETE</I> require much more handling.
complex query. A <I>Utility</I> query is processed by a
query-specific function in <A HREF="../../backend/commands">
commands.</A> A complex query, like <B>SELECT, UPDATE,</B> and
<B>DELETE</B> requires much more handling.
<P>
The parser takes the complex queries, and creates a
The parser takes a complex query, and creates a
<A HREF="../../include/nodes/parsenodes.h">Query</A> structure that
contains all the elements used by complex queries. Query.qual holds the
WHERE clause qualification, which is filled in by
<B>WHERE</B> clause qualification, which is filled in by
<A HREF="../../backend/parser/parse_clause.c">transformWhereClause().</A>
Each table referenced in the query is represented by a <A
HREF="../../include/nodes/parsenodes.h"> RangeTableEntry,</A> and they
@ -39,73 +39,70 @@ are linked together to form the <I>range table</I> of the query, which is
generated by <A HREF="../../backend/parser/parse_clause.c">
makeRangeTable().</A> Query.rtable holds the queries range table.
<P>
Certain queries, like SELECT, return columns of data. Other queries,
like INSERT and UPDATE, specify the columns modified by the query.
These column references are converted to <A
HREF="../../include/nodes/primnodes.h"> Resdom</A> entries, which are
Certain queries, like <B>SELECT,</B> return columns of data. Other
queries, like <B>INSERT</B> and <B>UPDATE,</B> specify the columns
modified by the query. These column references are converted to <A
HREF="../../include/nodes/primnodes.h">Resdom</A> entries, which are
linked together to make up the <I>target list</I> of the query. The
target list is stored in Query.targetList, which is generated by
<A HREF="../../backend/parser/parse_target.c">transformTargetList().</A>
<P>
Other query elements, like aggregates(SUM()), GROUP BY, ORDER BY are
also stored in their own Query fields.
Other query elements, like aggregates(<B>SUM()</B>), <B>GROUP BY,</B>
<B>ORDER BY</B> are also stored in their own Query fields.
<P>
The next step is for the Query to be modified by any VIEWS or RULES that
may apply to the query. This is performed by the <A
The next step is for the Query to be modified by any <B>VIEWS</B> or
<B>RULES</B> that may apply to the query. This is performed by the <A
HREF="../../backend/rewrite">rewrite</A> system.
<P>
The <A HREF="../../backend/optimizer">optimizer</A> takes the Query
structure, and generates an optimal
<A HREF="../..//include/nodes/plannodes.h">Plan</A> containing the
primitive operations to be performed by the executor to execute the
query. The <A HREF="../../backend/optimizer/path">path</A> module
determines the best table join order and join type of each table in the
RangeTable, using Query.qual(WHERE clause) to consider optimal index
usage.
structure and generates an optimal <A
HREF="../..//include/nodes/plannodes.h">Plan,</A> which contains the
operations to be performed to execute the query. The <A
HREF="../../backend/optimizer/path">path</A> module determines the best
table join order and join type of each table in the RangeTable, using
Query.qual(<B>WHERE</B> clause) to consider optimal index usage.
<P>
The Plan is then passed to the <A
HREF="../../backend/executor">executor</A> for execution, and the result
is returned to the client.
returned to the client.
<P>
There are many other modules that support this basic functionality.
There are many other modules that support this basic functionality.
They can be accessed by clicking on the flowchart.
<P>
Another area of interest is the shared memory area, containing
table data/index blocks, locks, and backend information:
<UL>
<LI>ShmemIndex - contains an index of all other shared memory
structures, allowing quick lookup of other structure locations in shared
memory
<LI>ShmemIndex - lookup of shared memory addresses using structure names
<LI><A HREF="../../include/storage/buf_internals.h">Buffer
Descriptors</A> - control header for shared memory buffer block
<LI><A HREF="../../include/storage/buf_internals.h">Buffer Blocks</A>
- block of table/index data shared by all backends
<LI>Shared Buf Lookup Table - lookup to see if a requested buffer
is already in the shared memory area
<LI><A HREF="../../include/storage/lock.h">LockTable</A>
- lock table structure, specifiying table, lock types, and
backends holding or waiting on lock
<LI>LockTable (lock hash) - lookup of LockTable structures using
table name
Descriptors</A> - control header for buffer cache block
<LI><A HREF="../../include/storage/buf_internals.h">Buffer Blocks</A> -
data/index buffer cache block
<LI>Shared Buf Lookup Table - lookup of buffer cache block address using
table name and block number(<A HREF="../../include/storage/buf_internals.h">
BufferTag</A>)
<LI><A HREF="../../include/storage/lock.h">LockTable (ctl)</A> - lock table
structure, specifiying table, lock types, and backends holding or
waiting on lock
<LI>LockTable (lock hash) - lookup of LockTable structures using relation
and database object ids
<LI>LockTable (xid hash) - lookup of LockTable structures using
transaction id
<LI><A HREF="../../include/storage/proc.h">Proc Header</A> - information
about each backend, including locks held/waiting, indexed by process id
</UL>
Each data structure is created by calling <A
HREF="../../backend/storage/ipc/shmem.c"> ShmemInitStruct(),</A> and
the lookup hashes are created by
HREF="../../backend/storage/ipc/shmem.c">ShmemInitStruct(),</A> and
the lookups are created by
<A HREF="../../backend/storage/ipc/shmem.c">ShmemInitHash().</A>
<HR>
<CENTER>
<EM><BIG>
Click on an item to see more detail or click
<A HREF="backend_dirs.html">here</a> to see the full index.
</EM></BIG>
Click on an item to see more detail or
<A HREF="backend_dirs.html">click</A> to see the full index.
</BIG></EM>
<BR>
<BR>
<IMG src="flow.jpg" usemap="#flowmap">
<IMG src="flow.jpg" usemap="#flowmap" alt="flowchart">
</CENTER>
<MAP name="flowmap">
<AREA COORDS="290,10,450,50" HREF="backend_dirs.html#main">