diff --git a/docs/reference/high-availability/cluster-design.asciidoc b/docs/reference/high-availability/cluster-design.asciidoc index 919e026bbeea..bddc8c206a7a 100644 --- a/docs/reference/high-availability/cluster-design.asciidoc +++ b/docs/reference/high-availability/cluster-design.asciidoc @@ -338,12 +338,16 @@ You should use <> to ensure that there is a copy of each shard in each zone. This means either zone remains fully available if the other zone fails. -All master-eligible nodes, including voting-only nodes, are on the critical path -for publishing cluster state updates. Because of this, these nodes require -reasonably fast persistent storage and a reliable, low-latency network -connection to the rest of the cluster. If you add a tiebreaker node in a third -independent zone then you must make sure it has adequate resources and good -connectivity to the rest of the cluster. +All master-eligible nodes, including voting-only nodes, are on the critical +path for <>. Cluster +state updates are usually independent of performance-critical workloads such as +indexing or searches, but they are involved in management activities such as +index creation and rollover, mapping updates, and recovery after a failure. The +performance characteristics of these activities are a function of the speed of +the storage on each master-eligible node, as well as the reliability and +latency of the network interconnections between all nodes in the cluster. You +must therefore ensure that the storage and networking available to the +nodes in your cluster are good enough to meet your performance goals. [[high-availability-cluster-design-three-zones]] ==== Clusters with three or more zones diff --git a/docs/reference/modules/discovery/publishing.asciidoc b/docs/reference/modules/discovery/publishing.asciidoc index 208386946d3f..af664585085c 100644 --- a/docs/reference/modules/discovery/publishing.asciidoc +++ b/docs/reference/modules/discovery/publishing.asciidoc @@ -1,38 +1,40 @@ [[cluster-state-publishing]] === Publishing the cluster state -The master node is the only node in a cluster that can make changes to the -cluster state. The master node processes one batch of cluster state updates at -a time, computing the required changes and publishing the updated cluster state -to all the other nodes in the cluster. Each publication starts with the master -broadcasting the updated cluster state to all nodes in the cluster. Each node -responds with an acknowledgement but does not yet apply the newly-received -state. Once the master has collected acknowledgements from enough -master-eligible nodes, the new cluster state is said to be _committed_ and the -master broadcasts another message instructing nodes to apply the now-committed -state. Each node receives this message, applies the updated state, and then -sends a second acknowledgement back to the master. +The elected master node is the only node in a cluster that can make changes to +the cluster state. The elected master node processes one batch of cluster state +updates at a time, computing the required changes and publishing the updated +cluster state to all the other nodes in the cluster. Each publication starts +with the elected master broadcasting the updated cluster state to all nodes in +the cluster. Each node responds with an acknowledgement but does not yet apply +the newly-received state. Once the elected master has collected +acknowledgements from enough master-eligible nodes, the new cluster state is +said to be _committed_ and the master broadcasts another message instructing +nodes to apply the now-committed state. Each node receives this message, +applies the updated state, and then sends a second acknowledgement back to the +master. -The master allows a limited amount of time for each cluster state update to be -completely published to all nodes. It is defined by the +The elected master allows a limited amount of time for each cluster state +update to be completely published to all nodes. It is defined by the `cluster.publish.timeout` setting, which defaults to `30s`, measured from the time the publication started. If this time is reached before the new cluster -state is committed then the cluster state change is rejected and the master -considers itself to have failed. It stands down and starts trying to elect a -new master. +state is committed then the cluster state change is rejected and the elected +master considers itself to have failed. It stands down and starts trying to +elect a new master node. If the new cluster state is committed before `cluster.publish.timeout` has -elapsed, the master node considers the change to have succeeded. It waits until -the timeout has elapsed or until it has received acknowledgements that each -node in the cluster has applied the updated state, and then starts processing -and publishing the next cluster state update. If some acknowledgements have not -been received (i.e. some nodes have not yet confirmed that they have applied -the current update), these nodes are said to be _lagging_ since their cluster -states have fallen behind the master's latest state. The master waits for the -lagging nodes to catch up for a further time, `cluster.follower_lag.timeout`, -which defaults to `90s`. If a node has still not successfully applied the -cluster state update within this time then it is considered to have failed and -is removed from the cluster. +elapsed, the elected master node considers the change to have succeeded. It +waits until the timeout has elapsed or until it has received acknowledgements +that each node in the cluster has applied the updated state, and then starts +processing and publishing the next cluster state update. If some +acknowledgements have not been received (i.e. some nodes have not yet confirmed +that they have applied the current update), these nodes are said to be +_lagging_ since their cluster states have fallen behind the elected master's +latest state. The elected master waits for the lagging nodes to catch up for a +further time, `cluster.follower_lag.timeout`, which defaults to `90s`. If a +node has still not successfully applied the cluster state update within this +time then it is considered to have failed and the elected master removes it +from the cluster. Cluster state updates are typically published as diffs to the previous cluster state, which reduces the time and network bandwidth needed to publish a cluster @@ -40,12 +42,19 @@ state update. For example, when updating the mappings for only a subset of the indices in the cluster state, only the updates for those indices need to be published to the nodes in the cluster, as long as those nodes have the previous cluster state. If a node is missing the previous cluster state, for example -when rejoining a cluster, the master will publish the full cluster state to -that node so that it can receive future updates as diffs. +when rejoining a cluster, the elected master will publish the full cluster +state to that node so that it can receive future updates as diffs. NOTE: {es} is a peer to peer based system, in which nodes communicate with one another directly. The high-throughput APIs (index, delete, search) do not -normally interact with the master node. The responsibility of the master node -is to maintain the global cluster state and reassign shards when nodes join or -leave the cluster. Each time the cluster state is changed, the new state is -published to all nodes in the cluster as described above. +normally interact with the elected master node. The responsibility of the +elected master node is to maintain the global cluster state which includes +reassigning shards when nodes join or leave the cluster. Each time the cluster +state is changed, the new state is published to all nodes in the cluster as +described above. + +The performance characteristics of cluster state updates are a function of the +speed of the storage on each master-eligible node, as well as the reliability +and latency of the network interconnections between all nodes in the cluster. +You must therefore ensure that the storage and networking available to the +nodes in your cluster are good enough to meet your performance goals. diff --git a/docs/reference/modules/node.asciidoc b/docs/reference/modules/node.asciidoc index b4a28b07db90..372af89af4a8 100644 --- a/docs/reference/modules/node.asciidoc +++ b/docs/reference/modules/node.asciidoc @@ -194,13 +194,6 @@ High availability (HA) clusters require at least three master-eligible nodes, at least two of which are not voting-only nodes. Such a cluster will be able to elect a master node even if one of the nodes fails. -Since voting-only nodes never act as the cluster's elected master, they may -require less heap and a less powerful CPU than the true master nodes. -However all master-eligible nodes, including voting-only nodes, require -reasonably fast persistent storage and a reliable and low-latency network -connection to the rest of the cluster, since they are on the critical path for -<>. - Voting-only master-eligible nodes may also fill other roles in your cluster. For instance, a node may be both a data node and a voting-only master-eligible node. A _dedicated_ voting-only master-eligible nodes is a voting-only @@ -212,6 +205,20 @@ dedicated voting-only master-eligible node, set: node.roles: [ master, voting_only ] ------------------- +Since dedicated voting-only nodes never act as the cluster's elected master, +they may require less heap and a less powerful CPU than the true master nodes. +However all master-eligible nodes, including voting-only nodes, are on the +critical path for <>. Cluster state updates are usually independent of +performance-critical workloads such as indexing or searches, but they are +involved in management activities such as index creation and rollover, mapping +updates, and recovery after a failure. The performance characteristics of these +activities are a function of the speed of the storage on each master-eligible +node, as well as the reliability and latency of the network interconnections +between the elected master node and the other nodes in the cluster. You must +therefore ensure that the storage and networking available to the nodes in your +cluster are good enough to meet your performance goals. + [[data-node]] ==== Data node