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
|
/*
* Licensed to Elasticsearch under one or more contributor
* license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright
* ownership. Elasticsearch licenses this file to you 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 org.elasticsearch.recovery;
import org.elasticsearch.action.admin.cluster.health.ClusterHealthRequestBuilder;
import org.elasticsearch.action.admin.cluster.health.ClusterHealthResponse;
import org.elasticsearch.action.admin.indices.recovery.RecoveryResponse;
import org.elasticsearch.cluster.ClusterState;
import org.elasticsearch.cluster.metadata.IndexMetaData;
import org.elasticsearch.cluster.routing.RecoverySource;
import org.elasticsearch.cluster.routing.UnassignedInfo;
import org.elasticsearch.common.Priority;
import org.elasticsearch.common.collect.MapBuilder;
import org.elasticsearch.common.settings.Settings;
import org.elasticsearch.common.unit.TimeValue;
import org.elasticsearch.discovery.zen.ZenDiscovery;
import org.elasticsearch.indices.recovery.RecoveryState;
import org.elasticsearch.test.ESIntegTestCase;
import org.elasticsearch.test.ESIntegTestCase.ClusterScope;
import org.elasticsearch.test.ESIntegTestCase.Scope;
import static org.elasticsearch.index.query.QueryBuilders.matchAllQuery;
import static org.elasticsearch.test.hamcrest.ElasticsearchAssertions.assertHitCount;
@ClusterScope(scope = Scope.TEST, numDataNodes = 0, transportClientRatio = 0.0)
public class FullRollingRestartIT extends ESIntegTestCase {
protected void assertTimeout(ClusterHealthRequestBuilder requestBuilder) {
ClusterHealthResponse clusterHealth = requestBuilder.get();
if (clusterHealth.isTimedOut()) {
logger.info("cluster health request timed out:\n{}", clusterHealth);
fail("cluster health request timed out");
}
}
@Override
protected int numberOfReplicas() {
return 1;
}
public void testFullRollingRestart() throws Exception {
Settings settings = Settings.builder().put(ZenDiscovery.JOIN_TIMEOUT_SETTING.getKey(), "30s").build();
internalCluster().startNode(settings);
createIndex("test");
final String healthTimeout = "1m";
for (int i = 0; i < 1000; i++) {
client().prepareIndex("test", "type1", Long.toString(i))
.setSource(MapBuilder.<String, Object>newMapBuilder().put("test", "value" + i).map()).execute().actionGet();
}
flush();
for (int i = 1000; i < 2000; i++) {
client().prepareIndex("test", "type1", Long.toString(i))
.setSource(MapBuilder.<String, Object>newMapBuilder().put("test", "value" + i).map()).execute().actionGet();
}
logger.info("--> now start adding nodes");
internalCluster().startNodesAsync(2, settings).get();
// make sure the cluster state is green, and all has been recovered
assertTimeout(client().admin().cluster().prepareHealth().setWaitForEvents(Priority.LANGUID).setTimeout(healthTimeout).setWaitForGreenStatus().setWaitForNoRelocatingShards(true).setWaitForNodes("3"));
logger.info("--> add two more nodes");
internalCluster().startNodesAsync(2, settings).get();
// We now have 5 nodes
setMinimumMasterNodes(3);
// make sure the cluster state is green, and all has been recovered
assertTimeout(client().admin().cluster().prepareHealth().setWaitForEvents(Priority.LANGUID).setTimeout(healthTimeout).setWaitForGreenStatus().setWaitForNoRelocatingShards(true).setWaitForNodes("5"));
logger.info("--> refreshing and checking data");
refresh();
for (int i = 0; i < 10; i++) {
assertHitCount(client().prepareSearch().setSize(0).setQuery(matchAllQuery()).get(), 2000L);
}
// now start shutting nodes down
internalCluster().stopRandomDataNode();
// make sure the cluster state is green, and all has been recovered
assertTimeout(client().admin().cluster().prepareHealth().setWaitForEvents(Priority.LANGUID).setTimeout(healthTimeout).setWaitForGreenStatus().setWaitForNoRelocatingShards(true).setWaitForNodes("4"));
// going down to 3 nodes. note that the min_master_node may not be in effect when we shutdown the 4th
// node, but that's OK as it is set to 3 before.
setMinimumMasterNodes(2);
internalCluster().stopRandomDataNode();
// make sure the cluster state is green, and all has been recovered
assertTimeout(client().admin().cluster().prepareHealth().setWaitForEvents(Priority.LANGUID).setTimeout(healthTimeout).setWaitForGreenStatus().setWaitForNoRelocatingShards(true).setWaitForNodes("3"));
logger.info("--> stopped two nodes, verifying data");
refresh();
for (int i = 0; i < 10; i++) {
assertHitCount(client().prepareSearch().setSize(0).setQuery(matchAllQuery()).get(), 2000L);
}
// closing the 3rd node
internalCluster().stopRandomDataNode();
// make sure the cluster state is green, and all has been recovered
assertTimeout(client().admin().cluster().prepareHealth().setWaitForEvents(Priority.LANGUID).setTimeout(healthTimeout).setWaitForGreenStatus().setWaitForNoRelocatingShards(true).setWaitForNodes("2"));
// closing the 2nd node
setMinimumMasterNodes(1);
internalCluster().stopRandomDataNode();
// make sure the cluster state is yellow, and all has been recovered
assertTimeout(client().admin().cluster().prepareHealth().setWaitForEvents(Priority.LANGUID).setTimeout(healthTimeout).setWaitForYellowStatus().setWaitForNoRelocatingShards(true).setWaitForNodes("1"));
logger.info("--> one node left, verifying data");
refresh();
for (int i = 0; i < 10; i++) {
assertHitCount(client().prepareSearch().setSize(0).setQuery(matchAllQuery()).get(), 2000L);
}
}
public void testNoRebalanceOnRollingRestart() throws Exception {
// see https://github.com/elastic/elasticsearch/issues/14387
internalCluster().startMasterOnlyNode(Settings.EMPTY);
internalCluster().startDataOnlyNodesAsync(3).get();
/**
* We start 3 nodes and a dedicated master. Restart on of the data-nodes and ensure that we got no relocations.
* Yet we have 6 shards 0 replica so that means if the restarting node comes back both other nodes are subject
* to relocating to the restarting node since all had 2 shards and now one node has nothing allocated.
* We have a fix for this to wait until we have allocated unallocated shards now so this shouldn't happen.
*/
prepareCreate("test").setSettings(Settings.builder().put(IndexMetaData.SETTING_NUMBER_OF_SHARDS, "6").put(IndexMetaData.SETTING_NUMBER_OF_REPLICAS, "0").put(UnassignedInfo.INDEX_DELAYED_NODE_LEFT_TIMEOUT_SETTING.getKey(), TimeValue.timeValueMinutes(1))).get();
for (int i = 0; i < 100; i++) {
client().prepareIndex("test", "type1", Long.toString(i))
.setSource(MapBuilder.<String, Object>newMapBuilder().put("test", "value" + i).map()).execute().actionGet();
}
ensureGreen();
ClusterState state = client().admin().cluster().prepareState().get().getState();
RecoveryResponse recoveryResponse = client().admin().indices().prepareRecoveries("test").get();
for (RecoveryState recoveryState : recoveryResponse.shardRecoveryStates().get("test")) {
assertTrue("relocated from: " + recoveryState.getSourceNode() + " to: " + recoveryState.getTargetNode() + "\n" + state,
recoveryState.getRecoverySource().getType() != RecoverySource.Type.PEER || recoveryState.getPrimary() == false);
}
internalCluster().restartRandomDataNode();
ensureGreen();
ClusterState afterState = client().admin().cluster().prepareState().get().getState();
recoveryResponse = client().admin().indices().prepareRecoveries("test").get();
for (RecoveryState recoveryState : recoveryResponse.shardRecoveryStates().get("test")) {
assertTrue("relocated from: " + recoveryState.getSourceNode() + " to: " + recoveryState.getTargetNode()+ "-- \nbefore: \n" + state,
recoveryState.getRecoverySource().getType() != RecoverySource.Type.PEER || recoveryState.getPrimary() == false);
}
}
}
|
