5 ms·

Fifty Years of Databases (2012)

45 points by yorp 12y ago

taspeotis 12y ago
IDS: The first NoSQL data store? In 1962 (sigmod.org) 1 point by yorp 3 minutes ago | flag | cached | share | discuss NoSQL doesn't appear anywhere on that page. The title should be Fifty Years of Databases (2012)
- dang 12y ago
  Right. We changed it back. Submitters: please don't editorialize in titles.
  - taspeotis 12y ago
    Thanks. I flag submissions with questionable titles. I believe this is the right thing to do, but what about after the title is changed? Does it still show up in a queue? Should I unflag submissions if I notice the title is changed?
    - dang 12y ago
      Sure, if the reason why you flagged a story gets fixed, unflagging it is a good idea.
- mariuz 12y ago
  So noSQL was absoleted by relational databases in '80
  - ArkyBeagle 12y ago
    It's hard to say, really. People were using a lot of non-SQL (read: row-based) databases well into the 1990s.
maxdemarzi 12y ago
>> In addition, IDS and other network systems encoded what we now think of as the “joins” between different kinds of records as part of the database structure rather than specifying them in each query. -- This is basically Neo4j and other graph databases. See: http://maxdemarzidotcom.files.wordpress.com/2012/08/screen-shot-2012-08-13-at-7-40-20-pm.png http://maxdemarzidotcom.files.wordpress.com/2012/08/screen-s... Those "R1", "R2"s in the picture... those are your "joins".
platz 12y ago
"the origins of the standard architecture for OLTP, which is row-based, and says it’s no longer optimal for any of the use cases that it is applied to. He proceeds to describe some research he has done, showing that row-based databases spend about 90 percent of their time acquiring and releasing locks, buffer management, and other activities that could be characterized as overhead in comparison to main task of reading and writing data. These results, which in Stonebraker’s view are intrinsic to the row-based architecture, require a new architecture to overcome. The discussion proceeds to a new database architecture, known as “NewSQL” or “NewOLTP,” which is single-threaded, lock-free, doesn’t require disk I/O in the critical path, and can scale out to a large multiple node cluster. Stonebraker criticizes the eventual consistency model that some NoSQL distributed systems employ and he defends the ACID guarantees as a superior model." http://www.se-radio.net/2013/12/episode-199-michael-stonebraker/ http://www.se-radio.net/2013/12/episode-199-michael-stonebra...
- ddorian43 12y ago
  he's describing voltdb (that he built)
  - ericfrenkiel 12y ago
    plus MemSQL, a YC company =)
    - _delirium 12y ago
      Two others are Google Spanner [1] and FoundationDB [2]. [1] see http://highscalability.com/blog/2012/9/24/google-spanners-most-surprising-revelation-nosql-is-out-and.html http://highscalability.com/blog/2012/9/24/google-spanners-mo... [2] https://foundationdb.com/acid-claims https://foundationdb.com/acid-claims

45 points by yorp 12y ago

IDS: The first NoSQL data store? In 1962 (sigmod.org) 1 point by yorp 3 minutes ago | flag | cached | share | discuss NoSQL doesn't appear anywhere on that page. The title should be Fifty Years of Databases (2012)

dang 12y ago
Right. We changed it back. Submitters: please don't editorialize in titles.
- taspeotis 12y ago
  Thanks. I flag submissions with questionable titles. I believe this is the right thing to do, but what about after the title is changed? Does it still show up in a queue? Should I unflag submissions if I notice the title is changed?
  - dang 12y ago
    Sure, if the reason why you flagged a story gets fixed, unflagging it is a good idea.
mariuz 12y ago
So noSQL was absoleted by relational databases in '80
- ArkyBeagle 12y ago
  It's hard to say, really. People were using a lot of non-SQL (read: row-based) databases well into the 1990s.

maxdemarzi 12y ago

>> In addition, IDS and other network systems encoded what we now think of as the “joins” between different kinds of records as part of the database structure rather than specifying them in each query. -- This is basically Neo4j and other graph databases. See: http://maxdemarzidotcom.files.wordpress.com/2012/08/screen-shot-2012-08-13-at-7-40-20-pm.png http://maxdemarzidotcom.files.wordpress.com/2012/08/screen-s... Those "R1", "R2"s in the picture... those are your "joins".

platz 12y ago

"the origins of the standard architecture for OLTP, which is row-based, and says it’s no longer optimal for any of the use cases that it is applied to. He proceeds to describe some research he has done, showing that row-based databases spend about 90 percent of their time acquiring and releasing locks, buffer management, and other activities that could be characterized as overhead in comparison to main task of reading and writing data. These results, which in Stonebraker’s view are intrinsic to the row-based architecture, require a new architecture to overcome. The discussion proceeds to a new database architecture, known as “NewSQL” or “NewOLTP,” which is single-threaded, lock-free, doesn’t require disk I/O in the critical path, and can scale out to a large multiple node cluster. Stonebraker criticizes the eventual consistency model that some NoSQL distributed systems employ and he defends the ACID guarantees as a superior model." http://www.se-radio.net/2013/12/episode-199-michael-stonebraker/ http://www.se-radio.net/2013/12/episode-199-michael-stonebra...

ddorian43 12y ago
he's describing voltdb (that he built)
- ericfrenkiel 12y ago
  plus MemSQL, a YC company =)
  - _delirium 12y ago
    Two others are Google Spanner [1] and FoundationDB [2]. [1] see http://highscalability.com/blog/2012/9/24/google-spanners-most-surprising-revelation-nosql-is-out-and.html http://highscalability.com/blog/2012/9/24/google-spanners-mo... [2] https://foundationdb.com/acid-claims https://foundationdb.com/acid-claims