README.md

simple-db

Code for all 6.830 labs will be available in this repo. Once you have set up your class repo, you pull lab code from here.

Directions for Repo Setup

Directions can be here

Lab Submission

Instructions for labs (including how to submit answers) are here

==========================================

simple DB structure:

8585347487

Type:

1. INT_TYPE
2. STRING_TYPE

TupleDesc:

1. TDItem 1. Type (fieldType) 2. String (fieldName)

RecordId (to a specific page of a specific table):

Field: 1. StringField 2. IntField

Tuple:

1. TupleDesc (schema)
2. RecordId 1. pageID 2. tupleNum
3. array of Fields (Field)

Catalog:

1. String (table name)
2. String (primary key)
3. DbFile 1. ID

HeapFile (one for each table):

1. set of HeapPage (each page has BufferPool.DEFAULT_PAGE_SIZE bytes)

HeapPage:

1. a set of slots (一个slot一个tuple, tuple有固定的大小 )
2. a header (bitmap) class fields:
   1. pageID
   2. tuple description
   3. header (bitmap)
   4. array of Tuples
   5. number of slots (maximum number of tuples)
   6. number of empty slots

Predicate: 1. 1 field (field ID that is the index of field in the Tuple object. Tuple里面的第几个field) 2. 1 operator (op) 3. 1 operand (Field)

"->" means "use"

HeapFile iterator -> Bufferpool.getPage() -> HeapFile readPage (read from disk)

HeapFile insert -> Bufferpool.getPage() -> HeapFile readPage

HeapFile delete -> Bufferpool.getPage() -> HeapFile readPage

Bufferpool insertTuple -> HeapFile insertTuple -> 1. HeapPage insert (on a free one or a new page) 2. writePage(). (write to disk)

Bufferpool deleteTuple -> HeapFile deleteTuple -> 1. HeapPage insert () 2. put page in the buffer pool

concurrency control

1. HeapFile.insertTuple(), HeapFile.deleteTuple() and HeapFile.iterator() should both use BufferPool.getPage()

2. BufferPool.insertTuple() and BufferPool.deleteTupe() call markDirty()

3. Release all locks associated with a transaction after it has committed or aborted to ensure strict 2PL