Skip to content

#2057 FetcherBolt: extract the crawl-delay policy #2205

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
rzo1 merged 1 commit into from
Oct 2, 2026
Merged

#2057 FetcherBolt: extract the crawl-delay policy #2205

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
114 changes: 114 additions & 0 deletions core/src/main/java/org/apache/stormcrawler/bolt/CrawlDelayPolicy.java
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,114 @@
/*
* 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.stormcrawler.bolt;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

/**
* Decides what the crawl delay of robots.txt does to the fetch queue of a URL, following
* fetcher.max.crawl.delay, fetcher.max.crawl.delay.force, fetcher.server.delay and
* fetcher.server.delay.force. Delays are in milliseconds.
*/
final class CrawlDelayPolicy {

// the bolt's category: log configurations for FetcherBolt keep covering these lines
private static final Logger LOG = LoggerFactory.getLogger(FetcherBolt.class);

enum Action {
/**
* Leave the fetch queue as it is: robots.txt has no delay, or the one the queue already
* has, even above the cap. The caller must not write back the delay it read, which could
* undo a concurrent raise.
*/
UNCHANGED,
/** Do not fetch the URL: robots.txt asks for a longer delay than accepted. */
SKIP,
APPLY
}

/**
* @param delay the delay to set on the fetch queue, for {@link Action#APPLY}
* @param robotsCrawlDelaySecs the robots.txt delay in seconds, rounded up, when it was longer
* than fetcher.max.crawl.delay and capped; null otherwise
*/
record Decision(Action action, long delay, String robotsCrawlDelaySecs) {
static final Decision UNCHANGED = new Decision(Action.UNCHANGED, 0, null);
static final Decision SKIP = new Decision(Action.SKIP, 0, null);
}

// max. delay accepted from robots.txt, negative for no limit
private final long maxCrawlDelay;
// whether maxCrawlDelay overwrites the longer value in robots.txt
// (otherwise URLs in this queue are skipped)
private final boolean maxCrawlDelayForce;
private final long serverDelay;
// whether the default delay is used even if the robots.txt
// specifies a shorter crawl-delay
private final boolean serverDelayForce;

CrawlDelayPolicy(
long maxCrawlDelay,
boolean maxCrawlDelayForce,
long serverDelay,
boolean serverDelayForce) {
this.maxCrawlDelay = maxCrawlDelay;
this.maxCrawlDelayForce = maxCrawlDelayForce;
this.serverDelay = serverDelay;
this.serverDelayForce = serverDelayForce;
}

/**
* @param url the URL, for the log
* @param queueId the ID of the fetch queue, for the log
* @param robotsDelay the crawl delay of robots.txt, not positive when there is none
*/
Decision decide(String url, String queueId, long robotsDelay, long queueDelay) {
if (robotsDelay <= 0 || robotsDelay == queueDelay) {
return Decision.UNCHANGED;
}
if (robotsDelay > maxCrawlDelay && maxCrawlDelay >= 0) {
if (!maxCrawlDelayForce) {
LOG.info("Crawl-Delay for {} too long ({}), skipping", url, robotsDelay);
return Decision.SKIP;
}
LOG.info(
"Crawl-Delay for {} too long ({}), using value of fetcher.max.crawl.delay"
+ " instead",
url,
robotsDelay);
// report the delay the fetcher is not holding, so a frontier-side
// consumer can enforce it at the source (#867)
return new Decision(
Action.APPLY, maxCrawlDelay, Long.toString(1L + ((robotsDelay - 1L) / 1000L)));
}
if (robotsDelay < serverDelay && serverDelayForce) {
LOG.info(
"Crawl delay for {} too short ({}), set to fetcher.server.delay",
url,
robotsDelay);
return new Decision(Action.APPLY, serverDelay, null);
}
LOG.info(
"Crawl delay for queue: {} is set to {} as per robots.txt. url: {}",
queueId,
robotsDelay,
url);
return new Decision(Action.APPLY, robotsDelay, null);
}
}
91 changes: 31 additions & 60 deletions core/src/main/java/org/apache/stormcrawler/bolt/FetcherBolt.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -103,6 +103,8 @@ public class FetcherBolt extends StatusEmitterBolt {

private RobotRulesLookup robotsLookup;

private CrawlDelayPolicy crawlDelayPolicy;

/** Largest number of helper threads ever alive; for tests. */
int helperPoolSize() {
return fetchHelpers == null ? 0 : fetchHelpers.largestPoolSize();
Expand All @@ -119,14 +121,6 @@ public Map<String, Object> getComponentConfiguration() {
/** This class picks items from queues and fetches the pages. */
private class FetcherThread extends Thread {

// max. delay accepted from robots.txt
private final long maxCrawlDelay;
// whether maxCrawlDelay overwrites the longer value in robots.txt
// (otherwise URLs in this queue are skipped)
private final boolean maxCrawlDelayForce;
// whether the default delay is used even if the robots.txt
// specifies a shorter crawl-delay
private final boolean crawlDelayForce;
private final int threadNum;

private long timeoutInQueues = -1;
Expand All @@ -138,10 +132,6 @@ public FetcherThread(Config conf, int num) {
this.setDaemon(true); // don't hang JVM on exit
this.setName("FetcherThread #" + num); // use an informative name

this.maxCrawlDelay = ConfUtils.getInt(conf, "fetcher.max.crawl.delay", 30) * 1000L;
this.maxCrawlDelayForce =
ConfUtils.getBoolean(conf, "fetcher.max.crawl.delay.force", false);
this.crawlDelayForce = ConfUtils.getBoolean(conf, "fetcher.server.delay.force", false);
this.threadNum = num;
timeoutInQueues = ConfUtils.getLong(conf, QUEUED_TIMEOUT_PARAM_KEY, timeoutInQueues);
protocolMetadataPrefix =
Expand Down Expand Up @@ -231,56 +221,30 @@ public void run() {
continue;
}
FetchItemQueue fiq = fetchQueues.getFetchItemQueue(fit.queueId, metadata);
if (rules.getCrawlDelay() > 0
&& rules.getCrawlDelay() != fiq.crawlDelay.get()) {
if (rules.getCrawlDelay() > maxCrawlDelay && maxCrawlDelay >= 0) {
boolean force = false;
String msg = "skipping";
if (maxCrawlDelayForce) {
force = true;
msg = "using value of fetcher.max.crawl.delay instead";
}
LOG.info(
"Crawl-Delay for {} too long ({}), {}",
CrawlDelayPolicy.Decision decision =
crawlDelayPolicy.decide(
fit.url,
rules.getCrawlDelay(),
msg);
if (force) {
fiq.crawlDelay.set(maxCrawlDelay);
// report the delay the fetcher is not holding, so a frontier-side
// consumer can enforce it at the source (#867)
robotsCrawlDelaySecs =
Long.toString(1L + ((rules.getCrawlDelay() - 1L) / 1000L));
metadata.setValue(
Constants.ROBOTS_CRAWL_DELAY_KEY, robotsCrawlDelaySecs);
} else {
// pass the info about crawl delay
metadata.setValue(Constants.STATUS_ERROR_CAUSE, "crawl_delay");
collector.emit(
org.apache.stormcrawler.Constants.StatusStreamName,
fit.tuple,
new Values(fit.url, metadata, Status.ERROR));
// no need to wait next time as we won't request
// from that site
asap = true;
continue;
}
} else if (rules.getCrawlDelay() < fetchQueues.crawlDelay
&& crawlDelayForce) {
fiq.crawlDelay.set(fetchQueues.crawlDelay);
LOG.info(
"Crawl delay for {} too short ({}), "
+ "set to fetcher.server.delay",
fit.url,
rules.getCrawlDelay());
} else {
fiq.crawlDelay.set(rules.getCrawlDelay());
LOG.info(
"Crawl delay for queue: {} is set to {} "
+ "as per robots.txt. url: {}",
fit.queueId,
fiq.crawlDelay.get(),
fit.url);
rules.getCrawlDelay(),
fiq.crawlDelay.get());
if (decision.action() == CrawlDelayPolicy.Action.SKIP) {
// pass the info about crawl delay
metadata.setValue(Constants.STATUS_ERROR_CAUSE, "crawl_delay");
collector.emit(
org.apache.stormcrawler.Constants.StatusStreamName,
fit.tuple,
new Values(fit.url, metadata, Status.ERROR));
// no need to wait next time as we won't request
// from that site
asap = true;
continue;
}
if (decision.action() == CrawlDelayPolicy.Action.APPLY) {
fiq.crawlDelay.set(decision.delay());
robotsCrawlDelaySecs = decision.robotsCrawlDelaySecs();
if (robotsCrawlDelaySecs != null) {
metadata.setValue(
Constants.ROBOTS_CRAWL_DELAY_KEY, robotsCrawlDelaySecs);
}
}

Expand Down Expand Up @@ -609,6 +573,13 @@ public void prepare(

this.fetchQueues = new FetchItemQueues(conf);

crawlDelayPolicy =
new CrawlDelayPolicy(
ConfUtils.getInt(conf, "fetcher.max.crawl.delay", 30) * 1000L,
ConfUtils.getBoolean(conf, "fetcher.max.crawl.delay.force", false),
fetchQueues.crawlDelay,
ConfUtils.getBoolean(conf, "fetcher.server.delay.force", false));

this.taskId = context.getThisTaskId();

int threadCount = ConfUtils.getInt(conf, "fetcher.threads.number", 10);
Expand Down
77 changes: 77 additions & 0 deletions core/src/test/java/org/apache/stormcrawler/bolt/CrawlDelayPolicyTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,77 @@
/*
* 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.stormcrawler.bolt;

import static org.junit.jupiter.api.Assertions.assertEquals;

import org.apache.stormcrawler.bolt.CrawlDelayPolicy.Action;
import org.apache.stormcrawler.bolt.CrawlDelayPolicy.Decision;
import org.junit.jupiter.params.ParameterizedTest;
import org.junit.jupiter.params.provider.CsvSource;

class CrawlDelayPolicyTest {

/**
* Delays in milliseconds. The columns: robots.txt delay, current delay of the fetch queue,
* fetcher.max.crawl.delay, fetcher.max.crawl.delay.force, fetcher.server.delay,
* fetcher.server.delay.force, then the expected decision: action, delay, robots crawl delay
* reported in seconds.
*/
@ParameterizedTest
@CsvSource(
textBlock =
"""
# no delay in robots.txt, or the one the fetch queue already has
-9223372036854775808, 1000, 30000, false, 1000, false, UNCHANGED, 0,
0, 1000, 30000, false, 1000, false, UNCHANGED, 0,
5000, 5000, 30000, false, 1000, false, UNCHANGED, 0,
60000, 60000, 30000, false, 1000, false, UNCHANGED, 0,
# longer than fetcher.max.crawl.delay, which applies before the server delay
60000, 1000, 30000, false, 1000, false, SKIP, 0,
60000, 1000, 30000, true, 1000, false, APPLY, 30000, 60
30500, 1000, 30000, true, 1000, false, APPLY, 30000, 31
60000, 1000, 30000, false, 1000, true, SKIP, 0,
60000, 1000, 30000, true, 1000, true, APPLY, 30000, 60
5000, 1000, 0, false, 1000, false, SKIP, 0,
# within fetcher.max.crawl.delay, or no limit when negative
30000, 1000, 30000, false, 1000, false, APPLY, 30000,
120000, 1000, -1, false, 1000, false, APPLY, 120000,
5000, 1000, 30000, false, 1000, true, APPLY, 5000,
# shorter than fetcher.server.delay
500, 0, 30000, false, 1000, true, APPLY, 1000,
500, 0, 30000, false, 1000, false, APPLY, 500,
""")
void decision(
long robotsDelay,
long queueDelay,
long maxCrawlDelay,
boolean maxCrawlDelayForce,
long serverDelay,
boolean serverDelayForce,
Action action,
long delay,
String reportedSecs) {
CrawlDelayPolicy policy =
new CrawlDelayPolicy(
maxCrawlDelay, maxCrawlDelayForce, serverDelay, serverDelayForce);

assertEquals(
new Decision(action, delay, reportedSecs),
policy.decide("http://a.test/", "a.test", robotsDelay, queueDelay));
}
}
Loading