A. DynamoDB - components

0. Table

• PK : partitionKey , or
• PK : partitionKey + Sortkey
• 
• Also, define mode for read/write operation:
• provisioned - define capacity RCU / WCU
• onDemand - uses RRU / WRU requests, internally. 2.5 times expensive.
• can switch b/w modes, at any time
• created : https://us-west-2.console.aws.amazon.com/dynamodbv2/home?region=us-west-2#table?name=ps-games

1. Record

• or Item
• has attributes (400 KB max)

2. Datatype

• Scalar Types – String, Number, Binary, Boolean, Null
• Document Types – List, Map
• Set Types – String Set, Number Set, Binary Set

3. TTL

• set expiration for record / item
• it will auto-delete and send event stream
• eg:
• enable TTL setting and give attribute name : my_expire_on
• add attribute : my_expire_on == <timeinMillisecond>
• expired after 2 hr
• and deleted from index - LSI and GSL
• and after 48 hrs, permanently deleted from main table also. (for recovery purpose)
• eg: webUser --> session 2 hr --> session logout --> expire his/her data after 2 hr.

4. Write Capacity Units (WCU)

• 1 WCU == write 1 item(upto 1 KB)/sec
• 

5. Read Capacity Units (RCU)

• 2 types of read
• ConsistentRead == True
  • 1 RCU == 1 Strongly Consistent Read of 1 item(upto 4 KB)
• ConsistentRead == false (default)
  • 1/2 RCU == Eventually Consistent Read of 1 item(upto 4 KB)
• 
• because of replication lag, can be Strongly or Eventually consistent
• 

---

6. Define throughput for each table

6.1. On-Demand Mode

• read/write operation, automatically scale up/down upto its max, with growing workloads
• Read Request Units (RRU)
• Write Request Units (WRU)
• for un-predictable workload.
• simplified billing but 2.5 times expensive

6.2. Provisioned Mode (default)

• for predicated workload
• can optionally, enable auto-scaling of WCU/RCU
• so we define capacity : RCU and WCU
• 
• 

6.3. ThrottleError

• if capacity exceeded then ProvisionedThroughputExceededException
• RCUs and WCUs are spread across all the table's partitions
• reason
• Hot Keys– one partition key is being read too many times (e.g., popular item)
• Hot Partitions
• Very large items, remember RCU and WCU depends on size of items
• Solutions:
• retry with Exponential backoff when exception is encountered (already in SDK)
• configure autoscale of WCU/RCU - [ min,max,desired ]
• Distribute partition keys as much as possible
• If RCU issue, we can use DAX

---

7. PartiQL

• SQL-compatible query language for DynamoDB - CRUD
• no joins
• Run PartiQL queries from:
• web Console
• NoSQL Workbench for DynamoDB
• CLI/SDK

---

8. Secondary Index

• udemy reference

8.1. LSI - Local Secondary Index

• index for : query by
• same PartitionKey
• Alternative SortKey
• restriction:
• 5 max
• Must be defined at table creation time
• Attribute Projections : KEYS_ONLY, INCLUDE, ALL
• Uses the WCUs and RCUs of the main table

8.2. GSI - Global Secondary Index

• index for : query by
• Alternative PartitionKey
• Alternative SortKey
• Speed up queries on non-key attributes
• Can be added/modified after table creation

• Attribute Projections : KEYS_ONLY, INCLUDE, ALL

• provision new WCU and RCU

• If the writes are throttled on the GSI,
• then the main table will be throttled!

---

9. transaction

• DynamoDB supports transactions but
• only for up to 25 items
• or 4 MB in total size
• ACID
• Read Modes (query/scan)
• Eventual Consistency 1/2 RCU
• Strong Consistency 1 RCU
• Transactional
  • Consumes 2x RCUs
  • TransactGetItems
• Write Modes ((add/update/delete))
• Standard
• Transactional
  • Consumes 2x WCUs
  • TransactWriteItems

---

10. store session states

• It’s common to use DynamoDB to store session states
• compare with other cache option/s
• vs. ElastiCache
  • ElastiCache is in-memory, but DynamoDB is serverless
  • Both are key/value stores
• vs. EFS
  • EFS must be attached to EC2 instances as a network drive
• vs. EBS & Instance Store
  • EBS & Instance Store can only be used for local caching, not shared caching
• vs. S3
  • S3 is higher latency, and not meant for small objects