331529fc94
* add an Aerospike storage backend * go mod vendor * add Aerospike storage configuration docs * review fixes * bump aerospike client to v3.1.1 * rename the defaultHostname variable * relocate the docs page
197 lines
5.4 KiB
Go
197 lines
5.4 KiB
Go
// Copyright 2013-2020 Aerospike, Inc.
|
|
//
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
// you may not use this file except in compliance with the License.
|
|
// You may obtain a copy of the License at
|
|
//
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
//
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
// See the License for the specific language governing permissions and
|
|
// limitations under the License.
|
|
|
|
package aerospike
|
|
|
|
import (
|
|
. "github.com/aerospike/aerospike-client-go/types"
|
|
)
|
|
|
|
type batchNode struct {
|
|
Node *Node
|
|
offsets []int
|
|
}
|
|
|
|
func newBatchNodeList(cluster *Cluster, policy *BatchPolicy, keys []*Key) ([]*batchNode, error) {
|
|
nodes := cluster.GetNodes()
|
|
|
|
if len(nodes) == 0 {
|
|
return nil, NewAerospikeError(SERVER_NOT_AVAILABLE, "command failed because cluster is empty.")
|
|
}
|
|
|
|
// Create initial key capacity for each node as average + 25%.
|
|
keysPerNode := len(keys) / len(nodes)
|
|
keysPerNode += keysPerNode / 2
|
|
|
|
// The minimum key capacity is 10.
|
|
if keysPerNode < 10 {
|
|
keysPerNode = 10
|
|
}
|
|
|
|
replicaPolicy := policy.ReplicaPolicy
|
|
replicaPolicySC := GetReplicaPolicySC(policy.GetBasePolicy())
|
|
|
|
// Split keys by server node.
|
|
batchNodes := make([]*batchNode, 0, len(nodes))
|
|
|
|
for i := range keys {
|
|
node, err := GetNodeBatchRead(cluster, keys[i], replicaPolicy, replicaPolicySC, 0, 0)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
if batchNode := findBatchNode(batchNodes, node); batchNode == nil {
|
|
batchNodes = append(batchNodes, newBatchNode(node, keysPerNode, i))
|
|
} else {
|
|
batchNode.AddKey(i)
|
|
}
|
|
}
|
|
return batchNodes, nil
|
|
}
|
|
|
|
func newBatchNodeListKeys(cluster *Cluster, policy *BatchPolicy, keys []*Key, sequenceAP, sequenceSC int, batchSeed *batchNode) ([]*batchNode, error) {
|
|
nodes := cluster.GetNodes()
|
|
|
|
if len(nodes) == 0 {
|
|
return nil, NewAerospikeError(SERVER_NOT_AVAILABLE, "command failed because cluster is empty.")
|
|
}
|
|
|
|
// Create initial key capacity for each node as average + 25%.
|
|
keysPerNode := len(keys) / len(nodes)
|
|
keysPerNode += keysPerNode / 2
|
|
|
|
// The minimum key capacity is 10.
|
|
if keysPerNode < 10 {
|
|
keysPerNode = 10
|
|
}
|
|
|
|
replicaPolicy := policy.ReplicaPolicy
|
|
replicaPolicySC := GetReplicaPolicySC(policy.GetBasePolicy())
|
|
|
|
// Split keys by server node.
|
|
batchNodes := make([]*batchNode, 0, len(nodes))
|
|
|
|
for _, offset := range batchSeed.offsets {
|
|
node, err := GetNodeBatchRead(cluster, keys[offset], replicaPolicy, replicaPolicySC, sequenceAP, sequenceSC)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
if batchNode := findBatchNode(batchNodes, node); batchNode == nil {
|
|
batchNodes = append(batchNodes, newBatchNode(node, keysPerNode, offset))
|
|
} else {
|
|
batchNode.AddKey(offset)
|
|
}
|
|
}
|
|
return batchNodes, nil
|
|
}
|
|
|
|
func newBatchNodeListRecords(cluster *Cluster, policy *BatchPolicy, records []*BatchRead, sequenceAP, sequenceSC int, batchSeed *batchNode) ([]*batchNode, error) {
|
|
nodes := cluster.GetNodes()
|
|
|
|
if len(nodes) == 0 {
|
|
return nil, NewAerospikeError(SERVER_NOT_AVAILABLE, "command failed because cluster is empty.")
|
|
}
|
|
|
|
// Create initial key capacity for each node as average + 25%.
|
|
keysPerNode := len(batchSeed.offsets) / len(nodes)
|
|
keysPerNode += keysPerNode / 2
|
|
|
|
// The minimum key capacity is 10.
|
|
if keysPerNode < 10 {
|
|
keysPerNode = 10
|
|
}
|
|
|
|
replicaPolicy := policy.ReplicaPolicy
|
|
replicaPolicySC := GetReplicaPolicySC(policy.GetBasePolicy())
|
|
|
|
// Split keys by server node.
|
|
batchNodes := make([]*batchNode, 0, len(nodes))
|
|
|
|
for _, offset := range batchSeed.offsets {
|
|
node, err := GetNodeBatchRead(cluster, records[offset].Key, replicaPolicy, replicaPolicySC, sequenceAP, sequenceSC)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
if batchNode := findBatchNode(batchNodes, node); batchNode == nil {
|
|
batchNodes = append(batchNodes, newBatchNode(node, keysPerNode, offset))
|
|
} else {
|
|
batchNode.AddKey(offset)
|
|
}
|
|
}
|
|
return batchNodes, nil
|
|
}
|
|
|
|
func newBatchIndexNodeList(cluster *Cluster, policy *BatchPolicy, records []*BatchRead) ([]*batchNode, error) {
|
|
nodes := cluster.GetNodes()
|
|
|
|
if len(nodes) == 0 {
|
|
return nil, NewAerospikeError(SERVER_NOT_AVAILABLE, "command failed because cluster is empty.")
|
|
}
|
|
|
|
// Create initial key capacity for each node as average + 25%.
|
|
keysPerNode := len(records) / len(nodes)
|
|
keysPerNode += keysPerNode / 2
|
|
|
|
// The minimum key capacity is 10.
|
|
if keysPerNode < 10 {
|
|
keysPerNode = 10
|
|
}
|
|
|
|
replicaPolicy := policy.ReplicaPolicy
|
|
replicaPolicySC := GetReplicaPolicySC(policy.GetBasePolicy())
|
|
|
|
// Split keys by server node.
|
|
batchNodes := make([]*batchNode, 0, len(nodes))
|
|
|
|
for i := range records {
|
|
node, err := GetNodeBatchRead(cluster, records[i].Key, replicaPolicy, replicaPolicySC, 0, 0)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
if batchNode := findBatchNode(batchNodes, node); batchNode == nil {
|
|
batchNodes = append(batchNodes, newBatchNode(node, keysPerNode, i))
|
|
} else {
|
|
batchNode.AddKey(i)
|
|
}
|
|
}
|
|
return batchNodes, nil
|
|
}
|
|
|
|
func newBatchNode(node *Node, capacity int, offset int) *batchNode {
|
|
res := &batchNode{
|
|
Node: node,
|
|
offsets: make([]int, 1, capacity),
|
|
}
|
|
|
|
res.offsets[0] = offset
|
|
return res
|
|
}
|
|
|
|
func (bn *batchNode) AddKey(offset int) {
|
|
bn.offsets = append(bn.offsets, offset)
|
|
}
|
|
|
|
func findBatchNode(nodes []*batchNode, node *Node) *batchNode {
|
|
for i := range nodes {
|
|
// Note: using pointer equality for performance.
|
|
if nodes[i].Node == node {
|
|
return nodes[i]
|
|
}
|
|
}
|
|
return nil
|
|
}
|