Please enable JavaScript.

Coggle requires JavaScript to display documents.

DynamoDB (Limits (Provisioned Throughput (Auto Scaling possible (Auto…

- - - - 1 read capacity unit = 1 strongly consistent read/sec
      - 1 read capacity unit = 2 eventually consistent read/sec
      - 1 write capacity unit = 1 write/sec for items
        
        all writes are max. 1 KB
      - read capacity chunk is 4KB
        
        round up to 4KB
    - - if > 10.000 writes or reads per second
    - - enabled by AWS Console (checkbox / capacity tab)
      - enabled by CLI or SDK
      - min. capacity (is 1 unit)
      - max. capacity
      - Auto Scaling policy
        
        can be changed anytime
        
        set to single table or index
        
        within a single region
  - - - must be UTF-8
      - min. 1 char up to 64KB / some 1-255 char
- - - - JSON
      - XML
      - HTML
  - - - accross multiple AZ in a region
- - - - UpdateTable
      - DeleteTable
      - CreateTable
    - - table size
      - state
      - index information
    - - list of all tables/account
  - - - max. 16 MB total retrieved size
      - retrieve max. 100 items
    - - new item
    - - write, replace or delete multiple items
    - - modfies attributes
    - - deletes item by primary key
    - - returns attributes
  - - - result is limited to 1 MB per call
        
        will be terminated if limit reached
        
        returns LastEvaluatedKey as start for the next scan
      - full scan across the table or secondary index
        
        slower than query
        
        as an iterator
    - - result is limited to 1 MB per call
      - returns 1 or more items by table PK
      - supports comparison operators
      - supports expressions
      - supports filter on non-key attributes
      - faster than scan
    - - ConsistentRead
        
        true = strongly consistency
        
        false = eventual consistency
  - - - Wrapper
      - interoperability between JS and DynamoDB datatypes
- - - - Time To Live (TTL)
      - contains a collection of attributes
      - primary key
        
        two types
        
        Partition key and sort key :check:
        
        composed of 2 attributes
        
        1st is partition key
        
        2nd is sort key
        
        partition key is not unique :warning:
        
        all items with the same partition key are stored togehter and sorted by sort key value
        
        but different sort key values
        
        sometimes sort key called range attribute also
        
        Partition key :check:
        
        Single Attribute
        
        sometimes called Hash Key (it is the same) :warning:
        
        partition key is unique
        
        like unique ID
        
        to identify items uniquely in a table
        
        dynamoDB uses pk value inside hash functions to get the partition (physical storage)
        
        partition key is called hash attribute also
        
        pk value
        
        must be scalar
        
        allowed data types
        
        string
        
        number
        
        binary
      - secondary indexes
        
        for more querying flexibility
        
        not required
        
        can read data in the same way as from the table
        
        two types
        
        Global secondary index
        
        different partition key from those on the table
        
        different sort key from those on the table
        
        Local secondary index
        
        same partition key as the table
        
        different sort key
    - - collection of data
      - similar to other db systems
      - contains zero to more items
      - items are like a rows or tuples
    - - a funcamental data element
      - like column data or fields
      - most attributes are scalar (only one value)
        
        string
        
        number
      - some have nested attributes
        
        up to 32 levels supported
    - - optional feature
      - captures data modification events
        
        near real time
        
        enabled on a table
        
        can trigger lambda funtions
      - events represented by a stream record
        
        write records for the following events
        
        new item added to table
        
        image of entire item, including all of its attributes
        
        item is updated
        
        "before" and "after" image of any modfied attribute
        
        item is deleted
        
        stream captures the image of the entire item before
        
        additionally contains
        
        table name
        
        timestamp
        
        other metadata
        
        stored for max. 24h