Mirror/MarginaliaSearch
1
0
Fork 0
mirror of https://github.com/MarginaliaSearch/MarginaliaSearch.git synced 2025-02-23 21:18:58 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Faster crawling

Browse Source
This commit is contained in:
vlofgren 2022-08-10 18:46:13 +02:00
parent ce09fce639
commit ffde8c8305
2 changed files with 33 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
marginalia_nu/src/main/java/nu/marginalia/wmsa/edge/crawling/CrawlerMain.java
View File

@ -29,11 +29,17 @@ public class CrawlerMain implements AutoCloseable {
new SynchronousQueue<>(), Util.threadFactory("OkHttp Dispatcher", true)));
private final UserAgent userAgent;
private final ThreadPoolExecutor pool;
final int poolSize = 256;
final int poolQueueSize = 32;
public CrawlerMain(EdgeCrawlPlan plan) throws Exception {
this.plan = plan;
this.userAgent = WmsaHome.getUserAgent();
BlockingQueue<Runnable> queue = new LinkedBlockingQueue<>(poolQueueSize);
pool = new ThreadPoolExecutor(poolSize/128, poolSize, 5, TimeUnit.MINUTES, queue); // maybe need to set -Xss for JVM to deal with this?
workLog = plan.createCrawlWorkLog();
crawlDataDir = plan.crawl.getDir();
}
@ -84,31 +90,44 @@ public class CrawlerMain implements AutoCloseable {
logger.info("Let's go");
final int poolSize = 1024;
BlockingQueue<Runnable> queue = new LinkedBlockingQueue<>(10);
ThreadPoolExecutor pool = new ThreadPoolExecutor(poolSize/128, poolSize, 5, TimeUnit.MINUTES, queue); // maybe need to set -Xss for JVM to deal with this?
AbortMonitor abortMonitor = AbortMonitor.getInstance();
Semaphore taskSem = new Semaphore(poolSize);
plan.forEachCrawlingSpecification(spec -> {
if (abortMonitor.isAlive()) {
pool.execute(() -> fetchDomain(spec));
try {
taskSem.acquire();
} catch (InterruptedException e) {
throw new RuntimeException(e);
}
pool.execute(() -> {
try {
fetchDomain(spec);
}
finally {
taskSem.release();
}
});
}
});
logger.info("Awaiting termination");
pool.shutdown();
while (!pool.awaitTermination(1, TimeUnit.SECONDS));
logger.info("All finished");
}
public void close() throws Exception {
logger.info("Awaiting termination");
pool.shutdown();
while (!pool.awaitTermination(1, TimeUnit.SECONDS));
logger.info("All finished");
workLog.close();
dispatcher.executorService().shutdownNow();
}
}

6
marginalia_nu/src/main/java/nu/marginalia/wmsa/edge/crawling/retreival/CrawlerRetreiver.java
View File

@ -83,8 +83,6 @@ public class CrawlerRetreiver {
}
public int fetch() throws IOException {
logger.info("Fetching {}", domain);
Optional<CrawledDomain> probeResult = probeDomainForProblems(domain);
if (probeResult.isPresent()) {
@ -272,10 +270,10 @@ public class CrawlerRetreiver {
@SneakyThrows
private void delay(long crawlDelay, long timeParsed) {
if (crawlDelay >= 1) {
if (timeParsed/1000 > crawlDelay)
if (timeParsed > crawlDelay)
return;
Thread.sleep(Math.min(1000*crawlDelay-timeParsed, 5000));
Thread.sleep(Math.min(crawlDelay-timeParsed, 5000));
}
else {
if (timeParsed > DEFAULT_CRAWL_DELAY_MS)