1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
|
use std::sync::Arc;
use garage_rpc::layout::*;
use garage_rpc::system::System;
use garage_util::data::*;
use crate::replication::*;
/// Sharded replication schema:
/// - based on the ring of nodes, a certain set of neighbors
/// store entries, given as a function of the position of the
/// entry's hash in the ring
/// - reads are done on all of the nodes that replicate the data
/// - writes as well
#[derive(Clone)]
pub struct TableShardedReplication {
/// The membership manager of this node
pub system: Arc<System>,
/// How many time each data should be replicated
pub replication_factor: usize,
/// How many nodes to contact for a read, should be at most `replication_factor`
pub read_quorum: usize,
/// How many nodes to contact for a write, should be at most `replication_factor`
pub write_quorum: usize,
}
impl TableReplication for TableShardedReplication {
type WriteSets = WriteLock<Vec<Vec<Uuid>>>;
fn storage_nodes(&self, hash: &Hash) -> Vec<Uuid> {
self.system.cluster_layout().storage_nodes_of(hash)
}
fn read_nodes(&self, hash: &Hash) -> Vec<Uuid> {
self.system.cluster_layout().read_nodes_of(hash)
}
fn read_quorum(&self) -> usize {
self.read_quorum
}
fn write_sets(&self, hash: &Hash) -> Self::WriteSets {
self.system.layout_manager.write_sets_of(hash)
}
fn write_quorum(&self) -> usize {
self.write_quorum
}
fn partition_of(&self, hash: &Hash) -> Partition {
self.system.cluster_layout().current().partition_of(hash)
}
fn sync_partitions(&self) -> SyncPartitions {
let layout = self.system.cluster_layout();
let layout_version = layout.ack_map_min();
let mut partitions = layout
.current()
.partitions()
.map(|(partition, first_hash)| {
let storage_sets = layout.storage_sets_of(&first_hash);
SyncPartition {
partition,
first_hash,
last_hash: [0u8; 32].into(), // filled in just after
storage_sets,
}
})
.collect::<Vec<_>>();
for i in 0..partitions.len() {
partitions[i].last_hash = if i + 1 < partitions.len() {
partitions[i + 1].first_hash
} else {
[0xFFu8; 32].into()
};
}
SyncPartitions {
layout_version,
partitions,
}
}
}
|
