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
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
|
// Copyright 2018 Yahoo Holdings. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
package com.yahoo.vespa.hosted.provision.provisioning;
import com.yahoo.config.provision.ApplicationId;
import com.yahoo.config.provision.ClusterSpec;
import com.yahoo.config.provision.NodeResources;
import com.yahoo.config.provision.NodeType;
import java.util.logging.Level;
import com.yahoo.vespa.hosted.provision.LockedNodeList;
import com.yahoo.vespa.hosted.provision.Node;
import com.yahoo.vespa.hosted.provision.NodeList;
import com.yahoo.vespa.hosted.provision.node.IP;
import com.yahoo.vespa.hosted.provision.persistence.NameResolver;
import java.util.EnumSet;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Optional;
import java.util.Set;
import java.util.function.Predicate;
import java.util.logging.Logger;
import java.util.stream.Collectors;
/**
* Builds up data structures necessary for node prioritization. It wraps each node
* up in a PrioritizableNode object with attributes used in sorting.
*
* The actual sorting/prioritization is implemented in the PrioritizableNode class as a compare method.
*
* @author smorgrav
*/
public class NodePrioritizer {
private final static Logger log = Logger.getLogger(NodePrioritizer.class.getName());
private final Map<Node, PrioritizableNode> nodes = new HashMap<>();
private final LockedNodeList allNodes;
private final DockerHostCapacity capacity;
private final NodeSpec requestedNodes;
private final ApplicationId application;
private final ClusterSpec clusterSpec;
private final NameResolver nameResolver;
private final boolean isDocker;
private final boolean isAllocatingForReplacement;
private final boolean isTopologyChange;
/** If set, a host can only have nodes by single tenant and does not allow in-place resizing. */
private final boolean allocateFully;
private final int currentClusterSize;
private final Set<Node> spareHosts;
NodePrioritizer(LockedNodeList allNodes, ApplicationId application, ClusterSpec clusterSpec, NodeSpec nodeSpec,
int spares, int wantedGroups, NameResolver nameResolver, HostResourcesCalculator hostResourcesCalculator,
boolean allocateFully) {
this.allNodes = allNodes;
this.capacity = new DockerHostCapacity(allNodes, hostResourcesCalculator);
this.requestedNodes = nodeSpec;
this.clusterSpec = clusterSpec;
this.application = application;
this.nameResolver = nameResolver;
this.spareHosts = findSpareHosts(allNodes, capacity, spares);
this.allocateFully = allocateFully;
NodeList nodesInCluster = allNodes.owner(application).type(clusterSpec.type()).cluster(clusterSpec.id());
NodeList nonRetiredNodesInCluster = nodesInCluster.not().retired();
long currentGroups = nonRetiredNodesInCluster.state(Node.State.active).stream()
.flatMap(node -> node.allocation()
.flatMap(alloc -> alloc.membership().cluster().group().map(ClusterSpec.Group::index))
.stream())
.distinct()
.count();
this.isTopologyChange = currentGroups != wantedGroups;
this.currentClusterSize = (int) nonRetiredNodesInCluster.state(Node.State.active).stream()
.map(node -> node.allocation().flatMap(alloc -> alloc.membership().cluster().group()))
.filter(clusterSpec.group()::equals)
.count();
this.isAllocatingForReplacement = isReplacement(
nodesInCluster.size(),
nodesInCluster.state(Node.State.failed).size());
this.isDocker = resources(requestedNodes) != null;
}
/**
* Spare hosts are the two hosts in the system with the most free capacity.
*
* We do not count retired or inactive nodes as used capacity (as they could have been
* moved to create space for the spare node in the first place).
*/
private static Set<Node> findSpareHosts(LockedNodeList nodes, DockerHostCapacity capacity, int spares) {
return nodes.asList().stream()
.filter(node -> node.type() == NodeType.host)
.filter(dockerHost -> dockerHost.state() == Node.State.active)
.filter(dockerHost -> capacity.freeIPs(dockerHost) > 0)
.sorted(capacity::compareWithoutInactive)
.limit(spares)
.collect(Collectors.toSet());
}
/** Returns the list of nodes sorted by PrioritizableNode::compare */
List<PrioritizableNode> prioritize() {
return nodes.values().stream().sorted().collect(Collectors.toList());
}
/**
* Add nodes that have been previously reserved to the same application from
* an earlier downsizing of a cluster
*/
void addSurplusNodes(List<Node> surplusNodes) {
for (Node node : surplusNodes) {
PrioritizableNode nodePri = toPrioritizable(node, true, false);
if (!nodePri.violatesSpares || isAllocatingForReplacement) {
nodes.put(node, nodePri);
}
}
}
/** Add a node on each docker host with enough capacity for the requested flavor */
void addNewDockerNodes(Predicate<Node> canAllocateTenantNodeTo) {
if ( ! isDocker) return;
LockedNodeList candidates = allNodes
.filter(node -> node.type() != NodeType.host || canAllocateTenantNodeTo.test(node))
.filter(node -> node.reservedTo().isEmpty() || node.reservedTo().get().equals(application.tenant()));
if (allocateFully) {
Set<String> candidateHostnames = candidates.asList().stream()
.filter(node -> node.type() == NodeType.tenant)
.filter(node -> node.allocation()
.map(a -> a.owner().tenant().equals(this.application.tenant()))
.orElse(false))
.flatMap(node -> node.parentHostname().stream())
.collect(Collectors.toSet());
candidates = candidates.filter(node -> candidateHostnames.contains(node.hostname()));
}
addNewDockerNodesOn(candidates);
}
private void addNewDockerNodesOn(LockedNodeList candidates) {
NodeResources wantedResources = resources(requestedNodes);
for (Node host : candidates) {
boolean hostHasCapacityForWantedFlavor = capacity.hasCapacity(host, wantedResources);
boolean conflictingCluster = allNodes.childrenOf(host).owner(application).asList().stream()
.anyMatch(child -> child.allocation().get().membership().cluster().id().equals(clusterSpec.id()));
if (!hostHasCapacityForWantedFlavor || conflictingCluster) continue;
log.log(Level.FINE, "Trying to add new Docker node on " + host);
Optional<IP.Allocation> allocation;
try {
allocation = host.ipConfig().pool().findAllocation(allNodes, nameResolver);
if (allocation.isEmpty()) continue; // No free addresses in this pool
} catch (Exception e) {
log.log(Level.WARNING, "Failed allocating IP address on " + host.hostname(), e);
continue;
}
Node newNode = Node.createDockerNode(allocation.get().addresses(),
allocation.get().hostname(),
host.hostname(),
resources(requestedNodes).with(host.flavor().resources().diskSpeed())
.with(host.flavor().resources().storageType()),
NodeType.tenant);
PrioritizableNode nodePri = toPrioritizable(newNode, false, true);
if ( ! nodePri.violatesSpares || isAllocatingForReplacement) {
log.log(Level.FINE, "Adding new Docker node " + newNode);
nodes.put(newNode, nodePri);
}
}
}
/** Add existing nodes allocated to the application */
void addApplicationNodes() {
EnumSet<Node.State> legalStates = EnumSet.of(Node.State.active, Node.State.inactive, Node.State.reserved);
allNodes.asList().stream()
.filter(node -> node.type() == requestedNodes.type())
.filter(node -> legalStates.contains(node.state()))
.filter(node -> node.allocation().isPresent())
.filter(node -> node.allocation().get().owner().equals(application))
.map(node -> toPrioritizable(node, false, false))
.forEach(prioritizableNode -> nodes.put(prioritizableNode.node, prioritizableNode));
}
/** Add nodes already provisioned, but not allocated to any application */
void addReadyNodes() {
allNodes.asList().stream()
.filter(node -> node.type() == requestedNodes.type())
.filter(node -> node.state() == Node.State.ready)
.map(node -> toPrioritizable(node, false, false))
.filter(n -> !n.violatesSpares || isAllocatingForReplacement)
.forEach(prioritizableNode -> nodes.put(prioritizableNode.node, prioritizableNode));
}
/**
* Convert a list of nodes to a list of node priorities. This includes finding, calculating
* parameters to the priority sorting procedure.
*/
private PrioritizableNode toPrioritizable(Node node, boolean isSurplusNode, boolean isNewNode) {
PrioritizableNode.Builder builder = new PrioritizableNode.Builder(node)
.surplusNode(isSurplusNode)
.newNode(isNewNode);
allNodes.parentOf(node).ifPresent(parent -> {
NodeResources parentCapacity = capacity.freeCapacityOf(parent, false);
builder.parent(parent).freeParentCapacity(parentCapacity);
if (!isNewNode)
builder.resizable(!allocateFully
&& requestedNodes.canResize(node.flavor().resources(), parentCapacity, isTopologyChange, currentClusterSize));
if (spareHosts.contains(parent))
builder.violatesSpares(true);
});
return builder.build();
}
private boolean isReplacement(int nofNodesInCluster, int nodeFailedNodes) {
if (nodeFailedNodes == 0) return false;
return requestedNodes.fulfilledBy(nofNodesInCluster - nodeFailedNodes);
}
private static NodeResources resources(NodeSpec requestedNodes) {
if ( ! (requestedNodes instanceof NodeSpec.CountNodeSpec)) return null;
return requestedNodes.resources().get();
}
}
|
