SOURSES
- NoSQL Summer Reding List
- NoSQL paper
Persistent Key/Value storage
High-performance transaction system applications typically insert rows in a History table to provide an activity trace; at the same time the transaction system generates log records for purposes of system recovery. Both types of generated information can benefit from efficient indexing.
- "QUOTE: The manner in which Bigtable uses memtables and SSTables to store updates to tablets is analogous to the way that the Log-Structured Merge Tree [ 26 ] stores updates to index data. In both systems, sorted data is buffered in memory before being written to disk, and reads must merge data from memory and disk."
- LSM-tree:
- "The Log-Structured Merge-Tree (LSM-Tree) [origin]"
- Fractal-tree, LSM-tree, B+tree (B+tree:InnoDB, LSM:Cassandra/BigTable, TokuDB:Fractal-tree)
- Sorted String Table (SSTable) or B+ Tree for a Database Index? (CouchDB:B+Tree, Cassandra:SSTable)
- "Performance Data For LevelDB, Berkley DB And BangDB For Random Operations"
- transactional Key/Value
- BTREE + HASH
- SSTable and MemTable
- Aries algorithm for WAL
- random read/write, sequential read and writes, batch operations
- compression
SQL
SQLite
MySQL(InnoDB)
MySQL(MyISAM)
Postgres
MS SQL
Oracle
DB2
Greenplum
NoSQL
GFS / HDFS
BigTable / HBase
MapRewduce / Hadoop
Cassandra
Dynamo
MongoDB
CouchDB
Riak
Megastore
Gmail, Picasa, Google Calendar, the Android Market and its AppEngine cloud all use Megastore
NewSQL
F1 /
Spanner (Paxos, TrueTime)
- wiki
- Spanner: Google’s Globally-Distributed Database
- Google reveals Spanner, the database tech that can span the planet
Distributed Shared Memory
Topics:
- UDP
- JGroups
- JXTA
- types:
- time:
- sync
- half-sync
- a-sync
- consistency
+ абсолютно консистентные (DSM)
- even cons
- кое-как (p2p)
- fault types
- memory consistency models
- DSM over message passing
- latency vs consistency dilemma
- lab: realize diff DSM over MP
- lab: what level for HttpSession replication?
- lab: business logic (relational model) over diff mem consistency models
- ?: примеры промышленных систем с DSM (Distr opensource caches, IMDG, Tomcat session replication)?
- lab: MMORPG
- lab: Bussines data in event consistent storage
- lab: realize even consistent over MP
Eventually consistent storage
- CAP theorem
- разные типы поломок
- eventually consistent systems: Amazon Dynamo/Cassandra/?(Erlang)
Group communications
- JGroups
- Gossip protocols
- distr collection info (p2p networks)
- membership protocols
- lab: realize p2p
Distributed commit
- types
- sync
- half-sync
- async
- FLP result
- 2PC
- XA spec, CORBA
- J2EE: JTA, JTS, OTS
- algorithm
- 3PC
- algorithm
- Paxos commit
- Paxos algorithm
- Paxos commit algorithm
Distributed time
Distributed Data Structures
Distributed Coordination
Queue solutions
- NoSQL Summer Reding List
- NoSQL paper
Persistent Key/Value storage
High-performance transaction system applications typically insert rows in a History table to provide an activity trace; at the same time the transaction system generates log records for purposes of system recovery. Both types of generated information can benefit from efficient indexing.
- "QUOTE: The manner in which Bigtable uses memtables and SSTables to store updates to tablets is analogous to the way that the Log-Structured Merge Tree [ 26 ] stores updates to index data. In both systems, sorted data is buffered in memory before being written to disk, and reads must merge data from memory and disk."
- LSM-tree:
- "The Log-Structured Merge-Tree (LSM-Tree) [origin]"
- Fractal-tree, LSM-tree, B+tree (B+tree:InnoDB, LSM:Cassandra/BigTable, TokuDB:Fractal-tree)
- Sorted String Table (SSTable) or B+ Tree for a Database Index? (CouchDB:B+Tree, Cassandra:SSTable)
- "Performance Data For LevelDB, Berkley DB And BangDB For Random Operations"
- transactional Key/Value
- BTREE + HASH
- SSTable and MemTable
- Aries algorithm for WAL
- random read/write, sequential read and writes, batch operations
- compression
SQL
SQLite
MySQL(InnoDB)
MySQL(MyISAM)
Postgres
MS SQL
Oracle
DB2
Greenplum
NoSQL
GFS / HDFS
BigTable / HBase
MapRewduce / Hadoop
Cassandra
Dynamo
MongoDB
CouchDB
Riak
Megastore
Gmail, Picasa, Google Calendar, the Android Market and its AppEngine cloud all use Megastore
NewSQL
F1 /
Spanner (Paxos, TrueTime)
- wiki
- Spanner: Google’s Globally-Distributed Database
- Google reveals Spanner, the database tech that can span the planet
Distributed Shared Memory
Topics:
- UDP
- JGroups
- JXTA
- types:
- time:
- sync
- half-sync
- a-sync
- consistency
+ абсолютно консистентные (DSM)
- even cons
- кое-как (p2p)
- fault types
- memory consistency models
- DSM over message passing
- latency vs consistency dilemma
- lab: realize diff DSM over MP
- lab: what level for HttpSession replication?
- lab: business logic (relational model) over diff mem consistency models
- ?: примеры промышленных систем с DSM (Distr opensource caches, IMDG, Tomcat session replication)?
- lab: MMORPG
- lab: Bussines data in event consistent storage
- lab: realize even consistent over MP
Eventually consistent storage
- CAP theorem
- разные типы поломок
- eventually consistent systems: Amazon Dynamo/Cassandra/?(Erlang)
Group communications
- JGroups
- Gossip protocols
- distr collection info (p2p networks)
- membership protocols
- lab: realize p2p
Distributed commit
- types
- sync
- half-sync
- async
- FLP result
- 2PC
- XA spec, CORBA
- J2EE: JTA, JTS, OTS
- algorithm
- 3PC
- algorithm
- Paxos commit
- Paxos algorithm
- Paxos commit algorithm
Distributed time
Distributed Data Structures
Distributed Coordination
Queue solutions
