ryanlecompte

6 ms·

1.
by ryanlecompte 14y ago
redis_failover provides a full automatic master/slave failover solution for Ruby. Changes in the 1.0 release: redis_failover now supports distributed monitoring among the Node Managers! Previously, the Node Managers were only used as a mean
2.
Redis_failover 1.0 Released - Distributed monitoring, failover strategies, more(github.com)
5 points by ryanlecompte 14y ago|1 comments
3.
by ryanlecompte 14y ago
Hey Steven, ZooKeeper is actually a proven technology for solving distributed configuration. It's widely used by Yahoo! and Netflix among other large companies. It's a proven PAXOS-like implementation, and definitely isn't a single point of
4.
by ryanlecompte 14y ago
FYI, redis_failover has now been rewritten to sit on top of ZooKeeper to deal with network partitions, stability, and data consistency. From the README: redis_failover attempts to provides a full automatic master/slave failover solution for
5.
by ryanlecompte 14y ago
That's right. The client still maintains direct connections with the actual master/slaves. It's only when it fails to connect with one of them that it goes to the failover daemon to ask for the current set of available nodes. The split of t
6.
by ryanlecompte 14y ago
The gem has a configurable --max-failures option that can be passed to failover daemon. The daemon will only mark a node as being unreachable if it fails to ping that amount of times (default 3). This might be something that can be improved
7.
by ryanlecompte 14y ago
From the README: Redis Failover attempts to provides a full automatic master/slave failover solution for Ruby. Redis does not provide an automatic failover capability when configured for master/slave replication. When the master node dies,
8.
Automatic Redis Failover for Ruby(github.com)
29 points by ryanlecompte 14y ago|11 comments