6 ms·

Totally! I can just sudo apt-get install low-latency-strong-consistency-multi-dc-with-declarative-query-language-db as a virtual package and get the right one f

Totally! I can just sudo apt-get install low-latency-strong-consistency-multi-dc-with-declarative-query-language-db as a virtual package and get the right one for my distro.

nickpsecurity 10y ago
Is that the command for installing Google's F1 RDBMS? It has those traits. I really wish they'd spin it off as a product since the only startup offering something similar got acquired and deep sixed by Apple.
- arielweisberg 10y ago
  I don't think it meets the low latency requirement. It's a CAP theorem joke. Can't have strong consistency across multiple data centers with low latency because you have to read/write from a quorum of DCs in order to maintain availability in case one DC goes down.
  - nickpsecurity 10y ago
    Oh yeah. Low latency eould be saying too much. I think its worst-case for strong-consistency is around 30 seconds or so. Anyway, closest to to low-latency part would be NonStop or OpenVMS clusers across redundant, leased lines. They're fast, scale well at HW level, use ACID databases, and high-availability. Commonly used in transaction processing. One VMS cluster survived 9/11 attack without losing a single transaction. NonStop does up to five 9's. Id be interested in the CAP analysis of these older methods.
    - arielweisberg 10y ago
      It also depends on your definition of data center. It's really multi-region that is the problem since then it's the speed of light you have to deal with. Quorum across some data centers is fast enough that it's not hard to manage in new applications.
      - nickpsecurity 10y ago
        That makes a lot of since. Many of the NonStop and VMS clusters were in same country but far enough from each other to isolate against many disasters. Could be why they did better on the issue.
  - mirekrusin 10y ago
    You can always use quantum entanglement to help a (qu)bit.
- lobster_johnson 10y ago
  Take a look at CockroachDB, which had its first public beta earlier this year [1]. It's directly inspired by Google's F1 and Spanner projects. It's similar to FoundationDB (the Apple product you're referring to) in that it's an SQL database layered on top of a K/V store. It's based on some clever technology to accomplish distributed transactions, strong consistency and high availability, and is looking very promising. [1] https://www.cockroachlabs.com/ https://www.cockroachlabs.com/
  - nickpsecurity 10y ago
    Appreciate the tip. Im aware of it. They had a recent post on HN where they were just getting around to dealing with stability issues. I was really excited till I saw that. Im holding off for a while till it matures a bit.
    - ddorian43 10y ago
      I think they just went too far with full-sql-joins-indexes-column-families in the first 1.0 version. Better to build it little by little.
  - carterehsmith 10y ago
    "First public beta" of a database product? No.
manigandham 10y ago
Great options already exist: MemSQL for blazing fast distributed full SQL database with cross-datacenter replication and in-memory rowstore + disk-based columnstore. ScyllaDB for Cassandra rewritten in C++ for blazing fast dynamo-style multi-master multi-datacenter disk-based wide-column database. I'll also throw in AMPS (by 60East) as the best messaging platform that supports innovative SQL and real-time state-of-the-world queries on it's message streams (instead of using that rabbitmq or kafka crap).
- akbar501 10y ago
  https://github.com/Netflix/dynomite https://github.com/Netflix/dynomite Dynomite for production proven, limitless scale for Redis. Provides ability to Redis for both in-memory and on-disk workloads. Dynamo-inspired linearly scalable, shared nothing multi-master architecture. Pairs well with Cassandra. Supports pluggable backends (ex. RocksDB). Been in production 2+ years. One of the larger clusters handles over 3.6 million sustained ops/sec in production, every day.
  - ddorian43 10y ago
    Redis is only the api/driver, not the features, right ? Meaning it's just crud and not the many features/data-types that redis provides, correct ?
    - akbar501 10y ago
      It's the full Redis API and protocol. See https://github.com/Netflix/dynomite/blob/dev/notes/redis.md https://github.com/Netflix/dynomite/blob/dev/notes/redis.md for a list of all supported Redis features. The benefit of Dynomite's support for the Redis API + protocol is a.) you can use any Redis client and b.) the same code can be used for standalone Redis on your laptop and on a distributed Dynomite cluster.
  - manigandham 10y ago
    This is neat, but why use this over just using Cassandra-based DBs? Also you were a speaker at Data Layer right?
biokoda 10y ago
http://www.actordb.com/ http://www.actordb.com/

Totally! I can just sudo apt-get install low-latency-strong-consistency-multi-dc-with-declarative-query-language-db as a virtual package and get the right one for my distro.

nickpsecurity 10y ago

Is that the command for installing Google's F1 RDBMS? It has those traits. I really wish they'd spin it off as a product since the only startup offering something similar got acquired and deep sixed by Apple.

arielweisberg 10y ago
I don't think it meets the low latency requirement. It's a CAP theorem joke. Can't have strong consistency across multiple data centers with low latency because you have to read/write from a quorum of DCs in order to maintain availability in case one DC goes down.
- nickpsecurity 10y ago
  Oh yeah. Low latency eould be saying too much. I think its worst-case for strong-consistency is around 30 seconds or so. Anyway, closest to to low-latency part would be NonStop or OpenVMS clusers across redundant, leased lines. They're fast, scale well at HW level, use ACID databases, and high-availability. Commonly used in transaction processing. One VMS cluster survived 9/11 attack without losing a single transaction. NonStop does up to five 9's. Id be interested in the CAP analysis of these older methods.
  - arielweisberg 10y ago
    It also depends on your definition of data center. It's really multi-region that is the problem since then it's the speed of light you have to deal with. Quorum across some data centers is fast enough that it's not hard to manage in new applications.
    - nickpsecurity 10y ago
      That makes a lot of since. Many of the NonStop and VMS clusters were in same country but far enough from each other to isolate against many disasters. Could be why they did better on the issue.
- mirekrusin 10y ago
  You can always use quantum entanglement to help a (qu)bit.
lobster_johnson 10y ago
Take a look at CockroachDB, which had its first public beta earlier this year [1]. It's directly inspired by Google's F1 and Spanner projects. It's similar to FoundationDB (the Apple product you're referring to) in that it's an SQL database layered on top of a K/V store. It's based on some clever technology to accomplish distributed transactions, strong consistency and high availability, and is looking very promising. [1] https://www.cockroachlabs.com/ https://www.cockroachlabs.com/
- nickpsecurity 10y ago
  Appreciate the tip. Im aware of it. They had a recent post on HN where they were just getting around to dealing with stability issues. I was really excited till I saw that. Im holding off for a while till it matures a bit.
  - ddorian43 10y ago
    I think they just went too far with full-sql-joins-indexes-column-families in the first 1.0 version. Better to build it little by little.
- carterehsmith 10y ago
  "First public beta" of a database product? No.

manigandham 10y ago

Great options already exist: MemSQL for blazing fast distributed full SQL database with cross-datacenter replication and in-memory rowstore + disk-based columnstore. ScyllaDB for Cassandra rewritten in C++ for blazing fast dynamo-style multi-master multi-datacenter disk-based wide-column database. I'll also throw in AMPS (by 60East) as the best messaging platform that supports innovative SQL and real-time state-of-the-world queries on it's message streams (instead of using that rabbitmq or kafka crap).

akbar501 10y ago
https://github.com/Netflix/dynomite https://github.com/Netflix/dynomite Dynomite for production proven, limitless scale for Redis. Provides ability to Redis for both in-memory and on-disk workloads. Dynamo-inspired linearly scalable, shared nothing multi-master architecture. Pairs well with Cassandra. Supports pluggable backends (ex. RocksDB). Been in production 2+ years. One of the larger clusters handles over 3.6 million sustained ops/sec in production, every day.
- ddorian43 10y ago
  Redis is only the api/driver, not the features, right ? Meaning it's just crud and not the many features/data-types that redis provides, correct ?
  - akbar501 10y ago
    It's the full Redis API and protocol. See https://github.com/Netflix/dynomite/blob/dev/notes/redis.md https://github.com/Netflix/dynomite/blob/dev/notes/redis.md for a list of all supported Redis features. The benefit of Dynomite's support for the Redis API + protocol is a.) you can use any Redis client and b.) the same code can be used for standalone Redis on your laptop and on a distributed Dynomite cluster.
- manigandham 10y ago
  This is neat, but why use this over just using Cassandra-based DBs? Also you were a speaker at Data Layer right?

biokoda 10y ago

http://www.actordb.com/ http://www.actordb.com/