github.com/yankunsam/loki/v2@v2.6.3-0.20220817130409-389df5235c27/docs/sources/fundamentals/architecture/rings.md (about) 1 --- 2 title: Consistent Hash Rings 3 weight: 40 4 --- 5 # Consistent Hash Rings 6 7 [Consistent hash rings](https://en.wikipedia.org/wiki/Consistent_hashing) 8 are incorporated into Loki cluster architectures to 9 10 - aid in the sharding of log lines 11 - implement high availability 12 - ease the horizontal scale up and scale down of clusters. 13 There is less of a performance hit for operations that must rebalance data. 14 15 Hash rings connect instances of a single type of component when 16 17 - there are a set of Loki instances in monolithic deployment mode 18 - there are multiple read components or multiple write components in 19 simple scalable deployment mode 20 - there are multiple instances of one type of component in microservices mode 21 22 Not all Loki components are connected by hash rings. 23 These components need to be connected into a hash ring: 24 25 - distributors 26 - ingesters 27 - query schedulers 28 - compactors 29 - rulers 30 31 These components can optionally be connected into a hash ring: 32 - index gateway 33 34 In an architecture that has three distributors and three ingestors defined, 35 the hash rings for these components connect the instances of same-type components. 36 37  38 39 Each node in the ring represents an instance of a component. 40 Each node has a key-value store that holds communication information 41 for each of the nodes in that ring. 42 Nodes update the key-value store periodically to keep the contents consistent 43 across all nodes. 44 For each node, the key-value store holds: 45 46 - an ID of the component node 47 - component address, used by other nodes as a communication channel 48 - an indication of the component node's health 49 50 ## Configuring rings 51 52 Define [ring configuration](../../../configuration/#ring_config) within the `common.ring_config` block. 53 54 Use the default `memberlist` key-value store type unless there is 55 a compelling reason to use a different key-value store type. 56 `memberlist` uses a [gossip protocol](https://en.wikipedia.org/wiki/Gossip_protocol) 57 to propagate information to all the nodes 58 to guarantee the eventual consistency of the key-value store contents. 59 60 There are additional configuration options for distributor rings, 61 ingester rings, and ruler rings. 62 These options are for advanced, specialized use only. 63 These options are defined within the `distributor.ring` block for distributors, 64 the `ingester.lifecycler.ring` block for ingesters, 65 and the `ruler.ring` block for rulers. 66 67 ## About the distributor ring 68 69 Distributors use the information in their key-value store 70 to keep a count of the quantity of distributors in the distributor ring. 71 The count further informs cluster limits. 72 73 ## About the ingester ring 74 75 Ingester ring information in the key-value stores is used by distributors. 76 The information lets the distributors shard log lines, 77 determining which ingester or set of ingesters a distributor sends log lines to. 78 79 ## About the query scheduler ring 80 81 Query schedulers use the information in their key-value store 82 for service discovery of the schedulers. 83 This allows queriers to connect to all available schedulers, 84 and it allows schedulers to connect to all available query frontends, 85 effectively creating a single queue that aids in balancing the query load. 86 87 ## About the compactor ring 88 89 Compactors use the information in the key-value store to identify 90 a single compactor instance that will be responsible for compaction. 91 The compactor is only enabled on the responsible instance, 92 despite the compactor target being on multiple instances. 93 94 ## About the ruler ring 95 96 The ruler ring is used to determine which rulers evaluate which rule groups. 97 98 ## About the index gateway ring 99 100 The index gateway ring is used to determine which gateway is responsible for which tenant's indexes when queried by rulers or queriers.