Cassandra 3.x High Availability(Second Edition)
上QQ阅读APP看书,第一时间看更新

Scaling consistency - the master-slave model

As distributed systems have become more commonplace, the need for higher capacity distributed databases has grown. Many distributed databases still attempt to maintain ACID guarantees (or in some cases only the consistency aspect, which is the most difficult in a distributed environment), leading to the master-slave architecture.

In this approach, there might be many servers handling requests, but only one server can actually perform writes so as to maintain data in a consistent state. This avoids the scenario where the same data can be modified via concurrent mutation requests to different nodes. The following diagram shows the most basic scenario:

However, we still have not solved the availability problem, as a failure of the write master would lead to application downtime. It also means that writes do not scale well, since they are all directed to a single machine.

Using sharding to scale writes

A variation on the master-slave approach that enables higher write volumes is a technique called sharding, in which the data is partitioned into groups of keys, such that one or more masters can own a known set of keys. For example, a database of user profiles can be partitioned by the last name, such that A-M belongs to one cluster and N-Z belongs to another, as follows:

An astute observer will notice that both master-slave and sharding introduce failure points on the master nodes, and in fact the sharding approach introduces multiple points of failure–one for each master! Additionally, the knowledge of where requests for certain keys go rests with the application layer, and adding shards requires manual shuffling of data to accommodate the modified key ranges.

Some systems employ shard managers as a layer of abstraction between the application and the physical shards. This has the effect of removing the requirement that the application must have knowledge of the partition map. However, it does not obviate the need for shuffling data as the cluster grows.

Handling the death of the leader

A common means of increasing availability in the event of a failure on a master node is to employ a master failover protocol. The particular semantics of the protocol vary among implementations, but the general principle is that a new master is appointed when the previous one fails. Not all failover algorithms are equal; however, in general, this feature increases availability in a master-slave system.

Even a master-slave database that employs leader election suffers from a number of undesirable traits:

  • Applications must understand the database topology
  • Data partitions must be carefully planned
  • Writes are difficult to scale
  • A failover dramatically increases the complexity of the system in general, and especially so for multisite databases
  • Adding capacity requires reshuffling data with a potential for downtime

Considering that our objective is a highly available system, and presuming that scalability is a concern, are there other options we need to consider?