SSDB is key-value(zset, hash, list) storage server, with Google LevelDB as storage engine, its goal is to replace Redis when encounter large sum of data that won’t fits the server’s memory.
SSDB Docs:

SSDB’s main features:

  • Uses Google LevelDB as storage engine
  • Master-slave replication, recover from break point after network failure
  • Online backup through network
  • Very suitable for storing collection data
  • Uses small RAM


$ssdb = new SimpleSSDB('', 8888);
$resp = $ssdb->set('key', '123');
$resp = $ssdb->get('key');
echo $resp; // output: 123


SSDB vs Redis

SSDB vs Redis

Views: 71126

52 Responses to "SSDB"

  • Hi,
    Can we look forward for further features like –
    1. getUnion(zset1, zset2, ..)
    2. getIntersection((zset1, zset2, ..)
    3. getUnionCount(zset1, zset2, ..) // approx using HLL
    4. getIntersectionCount(zset1, zset2, ..) // approx using HLL

    Is there any roadmap you have in mind ?

    Ashay Reply
  • Obviously Redis is single threaded and I was curious about whether methods like ‘FlushDB’ are also blocking in your library. I notice Redis have added an aysnc version of this method in version 4.

    Thanks and great library! Reply
  • Hi, excellent work! I was looking for something easier to wrap a java class to work with Basic4Java ( would I be able to use this as a cache server with cache expiry? Reply
  • Hi guys,

    Ssdb is now available as add-on for Heroku users. It’s currently in alpha stage.
    According to Heroku requirements, it needs to be tested by 10 alpha users.
    If you want to be one of those alpha users, leave your email on or just drop me a note to
    Alpha users will get a lifetime add-on plan of 200MB for free. To get account on Heroku is also free ;) Reply
  • Hi,

    This project is very interesting and there is a even ssdb-rocks. (

    The ssdb-rocks is bit old. Are there plans to provide support for latest version of rocksdb? Reply
  • Very interesting!
    But, i wonder. Whats the impact on the server?
    Are there any specifications for example:
    100.000 1Mb keys will use 800Mb memory, 1Gb disk space, 80ms to lookup a key.

    I don’t care about the disks pace. But I do care about the memory usage.
    Perhaps i’ll test it out sometime! Reply
    @Cagli: edit the ssdb.conf file, specify smaller cache_size and write_buffer_size, ssdb-server will use less memory. Reply
  • Hi!

    I want to use SSDB with CentOS 7 but it seems that it is not recommended:

    May I know the details? Can I use it after removing the limit on the maximum number of processes available to a single user (/etc/security/limits.d/90-nproc.conf) ? Reply
  • Hi, UI Download( ) use SSDB as the main cache to store vectors, icons, psds, and it works well, you can list my logo on your site: Reply

« [1][2][3][4] » 1/4

Leave a Comment