[
https://issues.apache.org/jira/browse/TINKERPOP-2245?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17319629#comment-17319629
]
ASF GitHub Bot commented on TINKERPOP-2245:
-------------------------------------------
divijvaidya commented on a change in pull request #1414:
URL: https://github.com/apache/tinkerpop/pull/1414#discussion_r611837557
##########
File path:
gremlin-server/src/main/java/org/apache/tinkerpop/gremlin/server/handler/MultiTaskSession.java
##########
@@ -0,0 +1,283 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF 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.apache.tinkerpop.gremlin.server.handler;
+
+import org.apache.tinkerpop.gremlin.driver.message.ResponseMessage;
+import org.apache.tinkerpop.gremlin.driver.message.ResponseStatusCode;
+import org.apache.tinkerpop.gremlin.groovy.engine.GremlinExecutor;
+import org.apache.tinkerpop.gremlin.groovy.jsr223.GroovyCompilerGremlinPlugin;
+import org.apache.tinkerpop.gremlin.jsr223.GremlinScriptEngine;
+import org.apache.tinkerpop.gremlin.jsr223.GremlinScriptEngineManager;
+import org.apache.tinkerpop.gremlin.server.Settings;
+import org.apache.tinkerpop.gremlin.structure.Transaction;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import javax.script.Bindings;
+import java.util.HashMap;
+import java.util.Map;
+import java.util.concurrent.BlockingQueue;
+import java.util.concurrent.ConcurrentMap;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.LinkedBlockingQueue;
+import java.util.concurrent.RejectedExecutionException;
+import java.util.concurrent.ScheduledExecutorService;
+import java.util.concurrent.ScheduledFuture;
+import java.util.concurrent.TimeUnit;
+import java.util.concurrent.atomic.AtomicBoolean;
+
+import static
org.apache.tinkerpop.gremlin.server.op.session.SessionOpProcessor.CONFIG_GLOBAL_FUNCTION_CACHE_ENABLED;
+
+/**
+ * A {@link Session} implementation that queues tasks given to it and executes
them in a serial fashion within the
+ * same thread which thus allows multiple tasks to be executed in the same
transaction. The first {@link SessionTask}
+ * to execute is supplied on the constructor and additional ones may be added
as they arrive with
+ * {@link #submitTask(SessionTask)} where they will be added to a queue where
they will await execution in the thread
+ * bound to this session.
+ */
+public class MultiTaskSession extends AbstractSession {
+ private static final Logger logger =
LoggerFactory.getLogger(MultiTaskSession.class);
+ protected final BlockingQueue<SessionTask> queue;
+ private final AtomicBoolean ending = new AtomicBoolean(false);
+ private final ScheduledExecutorService scheduledExecutorService;
+ private final GremlinScriptEngineManager scriptEngineManager;
+ private ScheduledFuture<?> requestCancelFuture;
+ private Bindings bindings;
+
+ /**
+ * Creates a new {@code MultiTaskSession} object providing the initial
starting {@link SessionTask} that gets
+ * executed by the session when it starts.
+ *
+ * @param initialSessionTask The tasks that starts the session.
+ * @param sessionId The id of the session
+ * @param sessions The session id to {@link Session} instances mapping
+ */
+ public MultiTaskSession(final SessionTask initialSessionTask, final String
sessionId,
+ final ConcurrentMap<String, Session> sessions) {
+ super(initialSessionTask, sessionId, false, sessions);
+
+ queue = new
LinkedBlockingQueue<>(initialSessionTask.getSettings().maxSessionTaskQueueSize);
+
+ // using a global function cache is cheaper than creating a new on per
session especially if you have to
+ // create a lot of sessions. it will generate a ton of throw-away
objects. mostly keeping the option open
+ // to not use it to preserve the ability to use the old functionality
if wanted or if there is some specific
+ // use case with sessions that needs it. if we wanted this could
eventually become a per-request option
+ // so that the client could control it as necessary and get
scriptengine isolation if they need it.
+ if (initialSessionTask.getSettings().useCommonEngineForSessions)
+ scriptEngineManager =
initialSessionTask.getGremlinExecutor().getScriptEngineManager();
+ else
+ scriptEngineManager =
initializeGremlinExecutor(initialSessionTask).getScriptEngineManager();
+
+ scheduledExecutorService =
initialSessionTask.getScheduledExecutorService();
+ submitTask(initialSessionTask);
+ }
+
+ /**
+ * Gets the script engine specific to this session which is dependent on
the
+ * {@link Settings#useCommonEngineForSessions} configuration.
+ */
+ @Override
+ public GremlinScriptEngine getScriptEngine(final SessionTask sessionTask,
final String language) {
+ return scriptEngineManager.getEngineByName(language);
+ }
+
+ @Override
+ public boolean isAcceptingTasks() {
+ return !ending.get();
+ }
+
+ @Override
+ public boolean submitTask(final SessionTask sessionTask) throws
RejectedExecutionException {
+ try {
+ return isAcceptingTasks() && queue.add(sessionTask);
+ } catch (IllegalStateException ise) {
+ final String msg = String.format("Task %s rejected from session
%s",
+ sessionTask.getRequestMessage().getRequestId(),
getSessionId());
+ throw new RejectedExecutionException(msg);
+ }
+ }
+
+ @Override
+ public void run() {
+ // allow the Session to know about the thread that is running it - the
thread really only has relevance
+ // once the session has started.
+ this.sessionThread = Thread.currentThread();
Review comment:
ah, in that case, if the default worker pool size is 8, then, we would
not be able to process 9th new concurrent session? Because all the threads will
be stuck serving the existing sessions? This will also impact single-session
aka session-less requests.
I understand that this is not different from current behaviour but today a
single session request does not get impacted by multi session query because
they share different worker thread pools. We are changing that assumption here.
A customer who is relying on the behaviour above to run longer/multiple queries
on sessions and low latency faster query on session-less will see a regression
from this change. We should document that somewhere.
In future, we should fix this limitation. This is the "starvation by session
requests" I was referring to in my email reply over the dev mailing list. There
are multiple ways to fix this but perhaps we should treat it as out of scope
for this PR.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
> Consolidate the executor for bytecode & string based client
> -----------------------------------------------------------
>
> Key: TINKERPOP-2245
> URL: https://issues.apache.org/jira/browse/TINKERPOP-2245
> Project: TinkerPop
> Issue Type: Improvement
> Components: server
> Affects Versions: 3.4.2
> Reporter: Divij Vaidya
> Assignee: Stephen Mallette
> Priority: Minor
>
> We have two code paths in the server which perform (more or less) the same
> functions. One is the executor for string based queries and other is the
> executor for bytecode. This code can be refactored together so that the logic
> to handle timeout, handle exception during execution, handle exception before
> execution and others can consolidated.
> [https://github.com/apache/tinkerpop/blob/master/gremlin-groovy/src/main/java/org/apache/tinkerpop/gremlin/groovy/engine/GremlinExecutor.java#L246]
> and
> [https://github.com/apache/tinkerpop/blob/master/gremlin-server/src/main/java/org/apache/tinkerpop/gremlin/server/op/traversal/TraversalOpProcessor.java#L333]
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
