private void removePublishTimeout() {
// to test that the acknowledgement mechanism is working we better disable the wait for publish
// otherwise the operation is most likely acknowledged even if it doesn't support ack
assertAcked(
client()
.admin()
.cluster()
.prepareUpdateSettings()
.setTransientSettings(
Settings.builder()
.put(DiscoverySettings.PUBLISH_TIMEOUT_SETTING.getKey(), "0")
.put(DiscoverySettings.COMMIT_TIMEOUT_SETTING.getKey(), "30s")));
}
/**
* Tests that cluster is committed or times out. It should never be the case that we fail an
* update due to a commit timeout, but it ends up being committed anyway
*/
public void testTimeoutOrCommit() throws Exception {
Settings settings =
Settings.builder()
// short but so we will sometime commit sometime timeout
.put(DiscoverySettings.COMMIT_TIMEOUT_SETTING.getKey(), "1ms")
.build();
MockNode master = createMockNode("master", settings, null);
MockNode node = createMockNode("node", settings, null);
ClusterState state =
ClusterState.builder(master.clusterState)
.nodes(
DiscoveryNodes.builder(master.clusterState.nodes())
.add(node.discoveryNode)
.masterNodeId(master.discoveryNode.getId()))
.build();
for (int i = 0; i < 10; i++) {
state = ClusterState.builder(state).incrementVersion().build();
logger.debug("--> publishing version [{}], UUID [{}]", state.version(), state.stateUUID());
boolean success;
try {
publishState(master.action, state, master.clusterState, 2).await(1, TimeUnit.HOURS);
success = true;
} catch (Discovery.FailedToCommitClusterStateException OK) {
success = false;
}
logger.debug("--> publishing [{}], verifying...", success ? "succeeded" : "failed");
if (success) {
assertSameState(node.clusterState, state);
} else {
assertThat(node.clusterState.stateUUID(), not(equalTo(state.stateUUID())));
}
}
}
public void testPublishingWithSendingErrors() throws Exception {
int goodNodes = randomIntBetween(2, 5);
int errorNodes = randomIntBetween(1, 5);
int timeOutNodes =
randomBoolean()
? 0
: randomIntBetween(1, 5); // adding timeout nodes will force timeout errors
final int numberOfMasterNodes = goodNodes + errorNodes + timeOutNodes + 1; // master
final boolean expectingToCommit = randomBoolean();
Settings.Builder settings = Settings.builder();
// make sure we have a reasonable timeout if we expect to timeout, o.w. one that will make the
// test "hang"
settings
.put(
DiscoverySettings.COMMIT_TIMEOUT_SETTING.getKey(),
expectingToCommit == false && timeOutNodes > 0 ? "100ms" : "1h")
.put(DiscoverySettings.PUBLISH_TIMEOUT_SETTING.getKey(), "5ms"); // test is about committing
MockNode master = createMockNode("master", settings.build(), null);
// randomize things a bit
int[] nodeTypes = new int[goodNodes + errorNodes + timeOutNodes];
for (int i = 0; i < goodNodes; i++) {
nodeTypes[i] = 0;
}
for (int i = goodNodes; i < goodNodes + errorNodes; i++) {
nodeTypes[i] = 1;
}
for (int i = goodNodes + errorNodes; i < nodeTypes.length; i++) {
nodeTypes[i] = 2;
}
Collections.shuffle(Arrays.asList(nodeTypes), random());
DiscoveryNodes.Builder discoveryNodesBuilder =
DiscoveryNodes.builder().add(master.discoveryNode);
for (int i = 0; i < nodeTypes.length; i++) {
final MockNode mockNode = createMockNode("node" + i);
discoveryNodesBuilder.add(mockNode.discoveryNode);
switch (nodeTypes[i]) {
case 1:
mockNode.action.errorOnSend.set(true);
break;
case 2:
mockNode.action.timeoutOnSend.set(true);
break;
}
}
final int dataNodes = randomIntBetween(0, 3); // data nodes don't matter
for (int i = 0; i < dataNodes; i++) {
final MockNode mockNode =
createMockNode(
"data_" + i,
Settings.builder().put(Node.NODE_MASTER_SETTING.getKey(), false).build(),
null);
discoveryNodesBuilder.add(mockNode.discoveryNode);
if (randomBoolean()) {
// we really don't care - just chaos monkey
mockNode.action.errorOnCommit.set(randomBoolean());
mockNode.action.errorOnSend.set(randomBoolean());
mockNode.action.timeoutOnCommit.set(randomBoolean());
mockNode.action.timeoutOnSend.set(randomBoolean());
}
}
final int minMasterNodes;
final String expectedBehavior;
if (expectingToCommit) {
minMasterNodes = randomIntBetween(0, goodNodes + 1); // count master
expectedBehavior = "succeed";
} else {
minMasterNodes = randomIntBetween(goodNodes + 2, numberOfMasterNodes); // +2 because of master
expectedBehavior = timeOutNodes > 0 ? "timeout" : "fail";
}
logger.info(
"--> expecting commit to {}. good nodes [{}], errors [{}], timeouts [{}]. min_master_nodes [{}]",
expectedBehavior,
goodNodes + 1,
errorNodes,
timeOutNodes,
minMasterNodes);
discoveryNodesBuilder
.localNodeId(master.discoveryNode.getId())
.masterNodeId(master.discoveryNode.getId());
DiscoveryNodes discoveryNodes = discoveryNodesBuilder.build();
MetaData metaData = MetaData.EMPTY_META_DATA;
ClusterState clusterState =
ClusterState.builder(CLUSTER_NAME).metaData(metaData).nodes(discoveryNodes).build();
ClusterState previousState = master.clusterState;
try {
publishState(master.action, clusterState, previousState, minMasterNodes);
if (expectingToCommit == false) {
fail("cluster state publishing didn't fail despite of not have enough nodes");
}
} catch (Discovery.FailedToCommitClusterStateException exception) {
logger.debug("failed to publish as expected", exception);
if (expectingToCommit) {
throw exception;
}
assertThat(exception.getMessage(), containsString(timeOutNodes > 0 ? "timed out" : "failed"));
}
}
