use of org.elasticsearch.transport.ConnectTransportException in project elasticsearch by elastic.
the class TransportInstanceSingleOperationActionTests method testRetryOfAnAlreadyTimedOutRequest.
public void testRetryOfAnAlreadyTimedOutRequest() throws Exception {
Request request = new Request().index("test").timeout(new TimeValue(0, TimeUnit.MILLISECONDS));
request.shardId = new ShardId("test", "_na_", 0);
PlainActionFuture<Response> listener = new PlainActionFuture<>();
setState(clusterService, ClusterStateCreationUtils.state("test", randomBoolean(), ShardRoutingState.STARTED));
action.new AsyncSingleAction(request, listener).start();
assertThat(transport.capturedRequests().length, equalTo(1));
long requestId = transport.capturedRequests()[0].requestId;
transport.clear();
DiscoveryNode node = clusterService.state().getNodes().getLocalNode();
transport.handleLocalError(requestId, new ConnectTransportException(node, "test exception"));
// wait until the timeout was triggered and we actually tried to send for the second time
assertBusy(new Runnable() {
@Override
public void run() {
assertThat(transport.capturedRequests().length, equalTo(1));
}
});
// let it fail the second time too
requestId = transport.capturedRequests()[0].requestId;
transport.handleLocalError(requestId, new ConnectTransportException(node, "test exception"));
try {
// result should return immediately
assertTrue(listener.isDone());
listener.get();
fail("this should fail with a transport exception");
} catch (ExecutionException t) {
if (ExceptionsHelper.unwrap(t, ConnectTransportException.class) == null) {
logger.info("expected ConnectTransportException but got ", t);
fail("expected and ConnectTransportException");
}
}
}
use of org.elasticsearch.transport.ConnectTransportException in project elasticsearch by elastic.
the class ExceptionSerializationTests method testConnectTransportException.
public void testConnectTransportException() throws IOException {
TransportAddress transportAddress = buildNewFakeTransportAddress();
DiscoveryNode node = new DiscoveryNode("thenode", transportAddress, emptyMap(), emptySet(), Version.CURRENT);
ConnectTransportException ex = serialize(new ConnectTransportException(node, "msg", "action", null));
assertEquals("[][" + transportAddress.toString() + "][action] msg", ex.getMessage());
assertEquals(node, ex.node());
assertEquals("action", ex.action());
assertNull(ex.getCause());
ex = serialize(new ConnectTransportException(node, "msg", "action", new NullPointerException()));
assertEquals("[][" + transportAddress + "][action] msg", ex.getMessage());
assertEquals(node, ex.node());
assertEquals("action", ex.action());
assertTrue(ex.getCause() instanceof NullPointerException);
}
use of org.elasticsearch.transport.ConnectTransportException in project elasticsearch by elastic.
the class MockTransportService method addUnresponsiveRule.
/**
* Adds a rule that will cause ignores each send request, simulating an unresponsive node
* and failing to connect once the rule was added.
*
* @param duration the amount of time to delay sending and connecting.
*/
public void addUnresponsiveRule(TransportAddress transportAddress, final TimeValue duration) {
final long startTime = System.currentTimeMillis();
addDelegate(transportAddress, new ClearableTransport(original) {
private final Queue<Runnable> requestsToSendWhenCleared = new LinkedBlockingDeque<Runnable>();
private boolean cleared = false;
TimeValue getDelay() {
return new TimeValue(duration.millis() - (System.currentTimeMillis() - startTime));
}
@Override
public void connectToNode(DiscoveryNode node, ConnectionProfile connectionProfile, CheckedBiConsumer<Connection, ConnectionProfile, IOException> connectionValidator) throws ConnectTransportException {
if (original.nodeConnected(node)) {
// connecting to an already connected node is a no-op
return;
}
TimeValue delay = getDelay();
if (delay.millis() <= 0) {
original.connectToNode(node, connectionProfile, connectionValidator);
return;
}
// TODO: Replace with proper setting
TimeValue connectingTimeout = NetworkService.TcpSettings.TCP_CONNECT_TIMEOUT.getDefault(Settings.EMPTY);
try {
if (delay.millis() < connectingTimeout.millis()) {
Thread.sleep(delay.millis());
original.connectToNode(node, connectionProfile, connectionValidator);
} else {
Thread.sleep(connectingTimeout.millis());
throw new ConnectTransportException(node, "UNRESPONSIVE: simulated");
}
} catch (InterruptedException e) {
throw new ConnectTransportException(node, "UNRESPONSIVE: simulated");
}
}
@Override
protected void sendRequest(Connection connection, long requestId, String action, TransportRequest request, TransportRequestOptions options) throws IOException {
// delayed sending - even if larger then the request timeout to simulated a potential late response from target node
TimeValue delay = getDelay();
if (delay.millis() <= 0) {
connection.sendRequest(requestId, action, request, options);
return;
}
// poor mans request cloning...
RequestHandlerRegistry reg = MockTransportService.this.getRequestHandler(action);
BytesStreamOutput bStream = new BytesStreamOutput();
request.writeTo(bStream);
final TransportRequest clonedRequest = reg.newRequest();
clonedRequest.readFrom(bStream.bytes().streamInput());
Runnable runnable = new AbstractRunnable() {
AtomicBoolean requestSent = new AtomicBoolean();
@Override
public void onFailure(Exception e) {
logger.debug("failed to send delayed request", e);
}
@Override
protected void doRun() throws IOException {
if (requestSent.compareAndSet(false, true)) {
connection.sendRequest(requestId, action, clonedRequest, options);
}
}
};
// store the request to send it once the rule is cleared.
synchronized (this) {
if (cleared) {
runnable.run();
} else {
requestsToSendWhenCleared.add(runnable);
threadPool.schedule(delay, ThreadPool.Names.GENERIC, runnable);
}
}
}
@Override
public void clearRule() {
synchronized (this) {
assert cleared == false;
cleared = true;
requestsToSendWhenCleared.forEach(Runnable::run);
}
}
});
}
use of org.elasticsearch.transport.ConnectTransportException in project crate by crate.
the class PeerRecoveryTargetService method doRecovery.
private void doRecovery(final long recoveryId) {
final StartRecoveryRequest request;
final RecoveryState.Timer timer;
CancellableThreads cancellableThreads;
try (RecoveryRef recoveryRef = onGoingRecoveries.getRecovery(recoveryId)) {
if (recoveryRef == null) {
LOGGER.trace("not running recovery with id [{}] - can not find it (probably finished)", recoveryId);
return;
}
final RecoveryTarget recoveryTarget = recoveryRef.target();
timer = recoveryTarget.state().getTimer();
cancellableThreads = recoveryTarget.cancellableThreads();
try {
assert recoveryTarget.sourceNode() != null : "can not do a recovery without a source node";
LOGGER.trace("{} preparing shard for peer recovery", recoveryTarget.shardId());
recoveryTarget.indexShard().prepareForIndexRecovery();
final long startingSeqNo = recoveryTarget.indexShard().recoverLocallyUpToGlobalCheckpoint();
assert startingSeqNo == UNASSIGNED_SEQ_NO || recoveryTarget.state().getStage() == RecoveryState.Stage.TRANSLOG : "unexpected recovery stage [" + recoveryTarget.state().getStage() + "] starting seqno [ " + startingSeqNo + "]";
request = getStartRecoveryRequest(LOGGER, clusterService.localNode(), recoveryTarget, startingSeqNo);
} catch (final Exception e) {
// this will be logged as warning later on...
LOGGER.trace("unexpected error while preparing shard for peer recovery, failing recovery", e);
onGoingRecoveries.failRecovery(recoveryId, new RecoveryFailedException(recoveryTarget.state(), "failed to prepare shard for recovery", e), true);
return;
}
}
Consumer<Exception> handleException = e -> {
if (LOGGER.isTraceEnabled()) {
LOGGER.trace(() -> new ParameterizedMessage("[{}][{}] Got exception on recovery", request.shardId().getIndex().getName(), request.shardId().id()), e);
}
Throwable cause = SQLExceptions.unwrap(e);
if (cause instanceof CancellableThreads.ExecutionCancelledException) {
// this can also come from the source wrapped in a RemoteTransportException
onGoingRecoveries.failRecovery(recoveryId, new RecoveryFailedException(request, "source has canceled the recovery", cause), false);
return;
}
if (cause instanceof RecoveryEngineException) {
// unwrap an exception that was thrown as part of the recovery
cause = cause.getCause();
}
// do it twice, in case we have double transport exception
cause = SQLExceptions.unwrap(cause);
if (cause instanceof RecoveryEngineException) {
// unwrap an exception that was thrown as part of the recovery
cause = cause.getCause();
}
if (cause instanceof IllegalIndexShardStateException || cause instanceof IndexNotFoundException || cause instanceof ShardNotFoundException) {
// if the target is not ready yet, retry
retryRecovery(recoveryId, "remote shard not ready", recoverySettings.retryDelayStateSync(), recoverySettings.activityTimeout());
return;
}
if (cause instanceof DelayRecoveryException) {
retryRecovery(recoveryId, cause, recoverySettings.retryDelayStateSync(), recoverySettings.activityTimeout());
return;
}
if (cause instanceof ConnectTransportException) {
LOGGER.debug("delaying recovery of {} for [{}] due to networking error [{}]", request.shardId(), recoverySettings.retryDelayNetwork(), cause.getMessage());
retryRecovery(recoveryId, cause.getMessage(), recoverySettings.retryDelayNetwork(), recoverySettings.activityTimeout());
return;
}
if (cause instanceof AlreadyClosedException) {
onGoingRecoveries.failRecovery(recoveryId, new RecoveryFailedException(request, "source shard is closed", cause), false);
return;
}
onGoingRecoveries.failRecovery(recoveryId, new RecoveryFailedException(request, e), true);
};
try {
LOGGER.trace("{} starting recovery from {}", request.shardId(), request.sourceNode());
cancellableThreads.executeIO(() -> transportService.sendRequest(request.sourceNode(), PeerRecoverySourceService.Actions.START_RECOVERY, request, new TransportResponseHandler<RecoveryResponse>() {
@Override
public void handleResponse(RecoveryResponse recoveryResponse) {
final TimeValue recoveryTime = new TimeValue(timer.time());
// do this through ongoing recoveries to remove it from the collection
onGoingRecoveries.markRecoveryAsDone(recoveryId);
if (LOGGER.isTraceEnabled()) {
StringBuilder sb = new StringBuilder();
sb.append('[').append(request.shardId().getIndex().getName()).append(']').append('[').append(request.shardId().id()).append("] ");
sb.append("recovery completed from ").append(request.sourceNode()).append(", took[").append(recoveryTime).append("]\n");
sb.append(" phase1: recovered_files [").append(recoveryResponse.phase1FileNames.size()).append("]").append(" with total_size of [").append(new ByteSizeValue(recoveryResponse.phase1TotalSize)).append("]").append(", took [").append(timeValueMillis(recoveryResponse.phase1Time)).append("], throttling_wait [").append(timeValueMillis(recoveryResponse.phase1ThrottlingWaitTime)).append(']').append("\n");
sb.append(" : reusing_files [").append(recoveryResponse.phase1ExistingFileNames.size()).append("] with total_size of [").append(new ByteSizeValue(recoveryResponse.phase1ExistingTotalSize)).append("]\n");
sb.append(" phase2: start took [").append(timeValueMillis(recoveryResponse.startTime)).append("]\n");
sb.append(" : recovered [").append(recoveryResponse.phase2Operations).append("]").append(" transaction log operations").append(", took [").append(timeValueMillis(recoveryResponse.phase2Time)).append("]").append("\n");
LOGGER.trace("{}", sb);
} else {
LOGGER.debug("{} recovery done from [{}], took [{}]", request.shardId(), request.sourceNode(), recoveryTime);
}
}
@Override
public void handleException(TransportException e) {
handleException.accept(e);
}
@Override
public String executor() {
// we do some heavy work like refreshes in the response so fork off to the generic threadpool
return ThreadPool.Names.GENERIC;
}
@Override
public RecoveryResponse read(StreamInput in) throws IOException {
return new RecoveryResponse(in);
}
}));
} catch (CancellableThreads.ExecutionCancelledException e) {
LOGGER.trace("recovery cancelled", e);
} catch (Exception e) {
handleException.accept(e);
}
}
use of org.elasticsearch.transport.ConnectTransportException in project crate by crate.
the class PreVoteCollectorTests method createObjects.
@Before
public void createObjects() {
Settings settings = Settings.builder().put(NODE_NAME_SETTING.getKey(), "node").build();
deterministicTaskQueue = new DeterministicTaskQueue(settings, random());
final MockTransport mockTransport = new MockTransport() {
@Override
protected void onSendRequest(final long requestId, final String action, final TransportRequest request, final DiscoveryNode node) {
super.onSendRequest(requestId, action, request, node);
assertThat(action, is(REQUEST_PRE_VOTE_ACTION_NAME));
assertThat(request, instanceOf(PreVoteRequest.class));
assertThat(node, not(equalTo(localNode)));
PreVoteRequest preVoteRequest = (PreVoteRequest) request;
assertThat(preVoteRequest.getSourceNode(), equalTo(localNode));
deterministicTaskQueue.scheduleNow(new Runnable() {
@Override
public void run() {
final PreVoteResponse response = responsesByNode.get(node);
if (response == null) {
handleRemoteError(requestId, new ConnectTransportException(node, "no response"));
} else {
handleResponse(requestId, response);
}
}
@Override
public String toString() {
return "response to " + request + " from " + node;
}
});
}
};
lastAcceptedTerm = randomNonNegativeLong();
currentTerm = randomLongBetween(lastAcceptedTerm, Long.MAX_VALUE);
lastAcceptedVersion = randomNonNegativeLong();
localNode = new DiscoveryNode("local-node", buildNewFakeTransportAddress(), Version.CURRENT);
responsesByNode.put(localNode, new PreVoteResponse(currentTerm, lastAcceptedTerm, lastAcceptedVersion));
transportService = mockTransport.createTransportService(settings, deterministicTaskQueue.getThreadPool(), boundTransportAddress -> localNode, null);
transportService.start();
transportService.acceptIncomingRequests();
preVoteCollector = new PreVoteCollector(transportService, () -> {
assert electionOccurred == false;
electionOccurred = true;
}, l -> {
});
// TODO need tests that check that the max term seen is updated
preVoteCollector.update(getLocalPreVoteResponse(), null);
}
Aggregations