(crawler) Make the use of virtual threads in the crawler configurable via system properties

This reverts commit 1c2426a052.

The change does not appear necessary to avoid problems.

build.gradle

@@ -43,12 +43,11 @@ subprojects.forEach {it ->
 }
 
 ext {
-    jvmVersion=23
-    dockerImageBase='container-registry.oracle.com/graalvm/jdk:23'
+    jvmVersion = 24
+    dockerImageBase='container-registry.oracle.com/graalvm/jdk:24'
     dockerImageTag='latest'
     dockerImageRegistry='marginalia'
     jibVersion = '3.4.4'
-
 }
 
 idea {

code/common/model/java/nu/marginalia/model/EdgeDomain.java

@@ -14,7 +14,7 @@ public class EdgeDomain implements Serializable {
     @Nonnull
     public final String topDomain;
 
-    public EdgeDomain(String host) {
+    public EdgeDomain(@Nonnull String host) {
         Objects.requireNonNull(host, "domain name must not be null");
 
         host = host.toLowerCase();
@@ -61,6 +61,10 @@ public class EdgeDomain implements Serializable {
         this.topDomain = topDomain;
     }
 
+    public static String getTopDomain(String host) {
+        return new EdgeDomain(host).topDomain;
+    }
+
     private boolean looksLikeGovTld(String host) {
         if (host.length() < 8)
             return false;
@@ -116,24 +120,6 @@ public class EdgeDomain implements Serializable {
         return topDomain.substring(0, cutPoint).toLowerCase();
     }
 
-    public String getLongDomainKey() {
-        StringBuilder ret = new StringBuilder();
-
-        int cutPoint = topDomain.indexOf('.');
-        if (cutPoint < 0) {
-            ret.append(topDomain);
-        } else {
-            ret.append(topDomain, 0, cutPoint);
-        }
-
-        if (!subDomain.isEmpty() && !"www".equals(subDomain)) {
-            ret.append(":");
-            ret.append(subDomain);
-        }
-
-        return ret.toString().toLowerCase();
-    }
-
     /** If possible, try to provide an alias domain,
      * i.e. a domain name that is very likely to link to this one
      * */

code/common/service/test/nu/marginalia/service/discovery/ZkServiceRegistryTest.java

@@ -25,7 +25,7 @@ import static org.mockito.Mockito.when;
 class ZkServiceRegistryTest {
     private static final int ZOOKEEPER_PORT = 2181;
     private static final GenericContainer<?> zookeeper =
-            new GenericContainer<>("zookeeper:3.8.0")
+            new GenericContainer<>("zookeeper:3.8")
                     .withExposedPorts(ZOOKEEPER_PORT);
 
     List<ZkServiceRegistry> registries = new ArrayList<>();

code/libraries/blocking-thread-pool/java/nu/marginalia/util/SimpleBlockingThreadPool.java

@@ -23,16 +23,33 @@ public class SimpleBlockingThreadPool {
     private final Logger logger = LoggerFactory.getLogger(SimpleBlockingThreadPool.class);
 
     public SimpleBlockingThreadPool(String name, int poolSize, int queueSize) {
+        this(name, poolSize, queueSize, ThreadType.PLATFORM);
+    }
+
+    public SimpleBlockingThreadPool(String name, int poolSize, int queueSize, ThreadType threadType) {
         tasks = new ArrayBlockingQueue<>(queueSize);
 
         for (int i = 0; i < poolSize; i++) {
-            Thread worker = new Thread(this::worker, name  + "[" + i + "]");
-            worker.setDaemon(true);
-            worker.start();
+
+            Thread.Builder threadBuilder = switch (threadType) {
+                case VIRTUAL -> Thread.ofVirtual();
+                case PLATFORM -> Thread.ofPlatform().daemon(true);
+            };
+
+            Thread worker = threadBuilder
+                    .name(name  + "[" + i + "]")
+                    .start(this::worker);
+
             workers.add(worker);
         }
 
     }
+
+    public enum ThreadType {
+        VIRTUAL,
+        PLATFORM
+    }
+
     public void submit(Task task) throws InterruptedException {
         tasks.put(task);
     }

code/processes/crawling-process/java/nu/marginalia/crawl/CrawlerMain.java

@@ -41,10 +41,7 @@ import java.nio.file.Files;
 import java.nio.file.Path;
 import java.nio.file.StandardCopyOption;
 import java.security.Security;
-import java.util.ArrayList;
-import java.util.Collections;
-import java.util.List;
-import java.util.Map;
+import java.util.*;
 import java.util.concurrent.ConcurrentHashMap;
 import java.util.concurrent.TimeUnit;
 import java.util.concurrent.atomic.AtomicInteger;
@@ -106,9 +103,18 @@ public class CrawlerMain extends ProcessMainClass {
         this.blacklist = blacklist;
         this.node = processConfiguration.node();
 
+        SimpleBlockingThreadPool.ThreadType threadType;
+        if (Boolean.getBoolean("crawler.useVirtualThreads")) {
+            threadType = SimpleBlockingThreadPool.ThreadType.VIRTUAL;
+        }
+        else {
+            threadType = SimpleBlockingThreadPool.ThreadType.PLATFORM;
+        }
+
         pool = new SimpleBlockingThreadPool("CrawlerPool",
                 Integer.getInteger("crawler.poolSize", 256),
-                1);
+                1,
+                threadType);
 
 
         // Wait for the blacklist to be loaded before starting the crawl
@@ -224,10 +230,7 @@ public class CrawlerMain extends ProcessMainClass {
 
         logger.info("Loaded {} domains", crawlSpecRecords.size());
 
-        // Shuffle the domains to ensure we get a good mix of domains in each crawl,
-        // so that e.g. the big domains don't get all crawled at once, or we end up
-        // crawling the same server in parallel from different subdomains...
-        Collections.shuffle(crawlSpecRecords);
+        crawlSpecRecords.sort(crawlSpecArrangement(crawlSpecRecords));
 
         // First a validation run to ensure the file is all good to parse
         if (crawlSpecRecords.isEmpty()) {
@@ -306,6 +309,30 @@ public class CrawlerMain extends ProcessMainClass {
         }
     }
 
+    /** Create a comparator that sorts the crawl specs in a way that is beneficial for the crawl,
+     * we want to enqueue domains that have common top domains first, but otherwise have a random
+     * order.
+     * <p></p>
+     * Note, we can't use hash codes for randomization as it is not desirable to have the same order
+     * every time the process is restarted (and CrawlSpecRecord is a record, which defines equals and
+     * hashcode based on the fields).
+     * */
+    private Comparator<CrawlSpecRecord> crawlSpecArrangement(List<CrawlSpecRecord> records) {
+        Random r = new Random();
+        Map<String, Integer> topDomainCounts = new HashMap<>(4 + (int) Math.sqrt(records.size()));
+        Map<String, Integer> randomOrder = new HashMap<>(records.size());
+
+        for (var spec : records) {
+            topDomainCounts.merge(EdgeDomain.getTopDomain(spec.domain), 1, Integer::sum);
+            randomOrder.put(spec.domain, r.nextInt());
+        }
+
+        return Comparator.comparing((CrawlSpecRecord spec) -> topDomainCounts.getOrDefault(EdgeDomain.getTopDomain(spec.domain), 0) >= 8)
+                .reversed()
+                .thenComparing(spec -> randomOrder.get(spec.domain))
+                .thenComparing(Record::hashCode); // non-deterministic tie-breaker to
+    }
+
     /** Submit a task for execution if it can be run, returns true if it was submitted
      * or if it can be discarded */
     private boolean trySubmitDeferredTask(CrawlTask task) {

code/processes/crawling-process/java/nu/marginalia/crawl/DomainStateDb.java

@@ -11,6 +11,7 @@ import java.nio.file.Path;
 import java.sql.Connection;
 import java.sql.DriverManager;
 import java.sql.SQLException;
+import java.time.Duration;
 import java.time.Instant;
 import java.util.Objects;
 import java.util.Optional;
@@ -24,6 +25,17 @@ public class DomainStateDb implements AutoCloseable {
 
     private final Connection connection;
 
+
+    public record CrawlMeta(
+            String domainName,
+            Instant lastFullCrawl,
+            Duration recrawlTime,
+            Duration crawlTime,
+            int recrawlErrors,
+            int crawlChanges,
+            int totalCrawlSize
+    ) {}
+
     public record SummaryRecord(
             String domainName,
             Instant lastUpdated,
@@ -102,6 +114,17 @@ public class DomainStateDb implements AutoCloseable {
                         feedUrl TEXT
                     )
                     """);
+            stmt.executeUpdate("""
+                    CREATE TABLE IF NOT EXISTS crawl_meta (
+                        domain TEXT PRIMARY KEY,
+                        lastFullCrawlEpochMs LONG NOT NULL,
+                        recrawlTimeMs LONG NOT NULL,
+                        recrawlErrors INTEGER NOT NULL,
+                        crawlTimeMs LONG NOT NULL,
+                        crawlChanges INTEGER NOT NULL,
+                        totalCrawlSize INTEGER NOT NULL
+                    )
+                    """);
             stmt.executeUpdate("""
                     CREATE TABLE IF NOT EXISTS favicon (
                         domain TEXT PRIMARY KEY,
@@ -164,6 +187,26 @@ public class DomainStateDb implements AutoCloseable {
         return Optional.empty();
     }
 
+    public void save(CrawlMeta crawlMeta) {
+        if (connection == null) throw new IllegalStateException("No connection to domainstate db");
+
+        try (var stmt = connection.prepareStatement("""
+                INSERT OR REPLACE INTO crawl_meta (domain, lastFullCrawlEpochMs, recrawlTimeMs, recrawlErrors, crawlTimeMs, crawlChanges, totalCrawlSize)
+                VALUES (?, ?, ?, ?, ?, ?, ?)
+                """)) {
+            stmt.setString(1, crawlMeta.domainName());
+            stmt.setLong(2, crawlMeta.lastFullCrawl.toEpochMilli());
+            stmt.setLong(3, crawlMeta.recrawlTime.toMillis());
+            stmt.setInt(4, crawlMeta.recrawlErrors);
+            stmt.setLong(5, crawlMeta.crawlTime.toMillis());
+            stmt.setInt(6, crawlMeta.crawlChanges);
+            stmt.setInt(7, crawlMeta.totalCrawlSize);
+            stmt.executeUpdate();
+        } catch (SQLException e) {
+            logger.error("Failed to insert crawl meta record", e);
+        }
+    }
+
     public void save(SummaryRecord record) {
         if (connection == null) throw new IllegalStateException("No connection to domainstate db");
 
@@ -182,7 +225,35 @@ public class DomainStateDb implements AutoCloseable {
         }
     }
 
-    public Optional<SummaryRecord> get(String domainName) {
+    public Optional<CrawlMeta> getMeta(String domainName) {
+        if (connection == null)
+            return Optional.empty();
+
+        try (var stmt = connection.prepareStatement("""
+                SELECT domain, lastFullCrawlEpochMs, recrawlTimeMs, recrawlErrors, crawlTimeMs, crawlChanges, totalCrawlSize
+                FROM crawl_meta
+                WHERE domain = ?
+                """)) {
+            stmt.setString(1, domainName);
+            var rs = stmt.executeQuery();
+            if (rs.next()) {
+                return Optional.of(new CrawlMeta(
+                        rs.getString("domain"),
+                        Instant.ofEpochMilli(rs.getLong("lastFullCrawlEpochMs")),
+                        Duration.ofMillis(rs.getLong("recrawlTimeMs")),
+                        Duration.ofMillis(rs.getLong("crawlTimeMs")),
+                        rs.getInt("recrawlErrors"),
+                        rs.getInt("crawlChanges"),
+                        rs.getInt("totalCrawlSize")
+                ));
+            }
+        } catch (SQLException ex) {
+            logger.error("Failed to get crawl meta record", ex);
+        }
+        return Optional.empty();
+    }
+
+    public Optional<SummaryRecord> getSummary(String domainName) {
         if (connection == null)
             return Optional.empty();
 

code/processes/crawling-process/java/nu/marginalia/crawl/fetcher/HttpFetcherImpl.java

@@ -29,6 +29,7 @@ import java.net.http.HttpResponse;
 import java.net.http.HttpTimeoutException;
 import java.time.Duration;
 import java.util.*;
+import java.util.concurrent.ExecutorService;
 import java.util.concurrent.Executors;
 import java.util.concurrent.Semaphore;
 import java.util.zip.GZIPInputStream;
@@ -56,12 +57,21 @@ public class HttpFetcherImpl implements HttpFetcher {
     private final HttpClient client;
 
     private HttpClient createClient() {
+        final ExecutorService executorService;
+
+        if (Boolean.getBoolean("crawler.httpclient.useVirtualThreads")) {
+            executorService = Executors.newVirtualThreadPerTaskExecutor();
+        }
+        else {
+            executorService = Executors.newCachedThreadPool();
+        }
+
         return HttpClient.newBuilder()
                 .sslContext(NoSecuritySSL.buildSslContext())
                 .cookieHandler(cookies)
                 .followRedirects(HttpClient.Redirect.NORMAL)
                 .connectTimeout(Duration.ofSeconds(8))
-                .executor(Executors.newCachedThreadPool())
+                .executor(executorService)
                 .build();
     }
 

code/processes/crawling-process/java/nu/marginalia/crawl/retreival/CrawlDelayTimer.java

@@ -3,6 +3,7 @@ package nu.marginalia.crawl.retreival;
 import nu.marginalia.crawl.fetcher.HttpFetcherImpl;
 
 import java.time.Duration;
+import java.util.concurrent.ThreadLocalRandom;
 
 import static java.lang.Math.max;
 import static java.lang.Math.min;
@@ -53,12 +54,13 @@ public class CrawlDelayTimer {
     public void waitFetchDelay(long spentTime) {
         long sleepTime = delayTime;
 
+        long jitter = ThreadLocalRandom.current().nextLong(0, 150);
         try {
             if (sleepTime >= 1) {
                 if (spentTime > sleepTime)
                     return;
 
-                Thread.sleep(min(sleepTime - spentTime, 5000));
+                Thread.sleep(min(sleepTime - spentTime, 5000) + jitter);
             } else {
                 // When no crawl delay is specified, lean toward twice the fetch+process time,
                 // within sane limits. This means slower servers get slower crawling, and faster
@@ -71,17 +73,17 @@ public class CrawlDelayTimer {
                 if (spentTime > sleepTime)
                     return;
 
-                Thread.sleep(sleepTime - spentTime);
+                Thread.sleep(sleepTime - spentTime + jitter);
             }
 
             if (slowDown) {
                 // Additional delay when the server is signalling it wants slower requests
-                Thread.sleep(DEFAULT_CRAWL_DELAY_MIN_MS);
+                Thread.sleep(DEFAULT_CRAWL_DELAY_MIN_MS + jitter);
             }
         }
         catch (InterruptedException e) {
             Thread.currentThread().interrupt();
-            throw new RuntimeException();
+            throw new RuntimeException("Interrupted", e);
         }
     }
 }

code/processes/crawling-process/java/nu/marginalia/crawl/retreival/CrawlerRetreiver.java

@@ -26,6 +26,8 @@ import java.io.IOException;
 import java.net.InetAddress;
 import java.net.UnknownHostException;
 import java.nio.file.Path;
+import java.time.Duration;
+import java.time.Instant;
 import java.util.List;
 import java.util.Optional;
 import java.util.concurrent.TimeUnit;
@@ -108,15 +110,24 @@ public class CrawlerRetreiver implements AutoCloseable {
                     DomainStateDb.SummaryRecord summaryRecord = sniffRootDocument(probedUrl, delayTimer);
                     domainStateDb.save(summaryRecord);
 
+                    if (Thread.interrupted()) {
+                        // There's a small chance we're interrupted during the sniffing portion
+                        throw new InterruptedException();
+                    }
+
+                    Instant recrawlStart = Instant.now();
+                    CrawlerRevisitor.RecrawlMetadata recrawlMetadata = crawlerRevisitor.recrawl(oldCrawlData, robotsRules, delayTimer);
+                    Duration recrawlTime = Duration.between(recrawlStart, Instant.now());
+
                     // Play back the old crawl data (if present) and fetch the documents comparing etags and last-modified
-                    if (crawlerRevisitor.recrawl(oldCrawlData, robotsRules, delayTimer) > 0) {
+                    if (recrawlMetadata.size() > 0) {
                         // If we have reference data, we will always grow the crawl depth a bit
                         crawlFrontier.increaseDepth(1.5, 2500);
                     }
 
                     oldCrawlData.close(); // proactively close the crawl data reference here to not hold onto expensive resources
 
-                    yield crawlDomain(probedUrl, robotsRules, delayTimer, domainLinks);
+                    yield crawlDomain(probedUrl, robotsRules, delayTimer, domainLinks, recrawlMetadata, recrawlTime);
                 }
                 case HttpFetcher.DomainProbeResult.Redirect(EdgeDomain domain1) -> {
                     domainStateDb.save(DomainStateDb.SummaryRecord.forError(domain, "Redirect", domain1.toString()));
@@ -138,8 +149,11 @@ public class CrawlerRetreiver implements AutoCloseable {
     private int crawlDomain(EdgeUrl rootUrl,
                             SimpleRobotRules robotsRules,
                             CrawlDelayTimer delayTimer,
-                            DomainLinks domainLinks) {
+                            DomainLinks domainLinks,
+                            CrawlerRevisitor.RecrawlMetadata recrawlMetadata,
+                            Duration recrawlTime) {
 
+        Instant crawlStart = Instant.now();
 
         // Add external links to the crawl frontier
         crawlFrontier.addAllToQueue(domainLinks.getUrls(rootUrl.proto));
@@ -149,6 +163,8 @@ public class CrawlerRetreiver implements AutoCloseable {
             crawlFrontier.addAllToQueue(fetcher.fetchSitemapUrls(sitemap, delayTimer));
         }
 
+        int crawlerAdditions = 0;
+
         while (!crawlFrontier.isEmpty()
             && !crawlFrontier.isCrawlDepthReached()
             && errorCount < MAX_ERRORS
@@ -180,7 +196,11 @@ public class CrawlerRetreiver implements AutoCloseable {
                 continue;
 
             try {
-                fetchContentWithReference(top, delayTimer, DocumentWithReference.empty());
+                var result = fetchContentWithReference(top, delayTimer, DocumentWithReference.empty());
+
+                if (result.isOk()) {
+                    crawlerAdditions++;
+                }
             }
             catch (InterruptedException ex) {
                 Thread.currentThread().interrupt();
@@ -188,6 +208,17 @@ public class CrawlerRetreiver implements AutoCloseable {
             }
         }
 
+        Duration crawlTime = Duration.between(crawlStart, Instant.now());
+        domainStateDb.save(new DomainStateDb.CrawlMeta(
+                domain,
+                Instant.now(),
+                recrawlTime,
+                crawlTime,
+                recrawlMetadata.errors(),
+                crawlerAdditions,
+                recrawlMetadata.size() + crawlerAdditions
+        ));
+
         return crawlFrontier.visitedSize();
     }
 
@@ -289,6 +320,10 @@ public class CrawlerRetreiver implements AutoCloseable {
         }
         catch (Exception ex) {
             logger.error("Error configuring link filter", ex);
+            if (Thread.interrupted()) {
+                Thread.currentThread().interrupt();
+                return DomainStateDb.SummaryRecord.forError(domain, "Crawler Interrupted", ex.getMessage());
+            }
         }
         finally {
             crawlFrontier.addVisited(rootUrl);
@@ -316,7 +351,7 @@ public class CrawlerRetreiver implements AutoCloseable {
     );
 
     private Optional<String> guessFeedUrl(CrawlDelayTimer timer) throws InterruptedException {
-        var oldDomainStateRecord = domainStateDb.get(domain);
+        var oldDomainStateRecord = domainStateDb.getSummary(domain);
 
         // If we are already aware of an old feed URL, then we can just revalidate it
         if (oldDomainStateRecord.isPresent()) {

code/processes/crawling-process/java/nu/marginalia/crawl/retreival/revisit/CrawlerRevisitor.java

@@ -31,7 +31,7 @@ public class CrawlerRevisitor {
     }
 
     /** Performs a re-crawl of old documents, comparing etags and last-modified */
-    public int recrawl(CrawlDataReference oldCrawlData,
+    public RecrawlMetadata recrawl(CrawlDataReference oldCrawlData,
                        SimpleRobotRules robotsRules,
                        CrawlDelayTimer delayTimer)
     throws InterruptedException {
@@ -39,6 +39,7 @@ public class CrawlerRevisitor {
         int retained = 0;
         int errors = 0;
         int skipped = 0;
+        int size = 0;
 
         for (CrawledDocument doc : oldCrawlData) {
             if (errors > 20) {
@@ -46,6 +47,10 @@ public class CrawlerRevisitor {
                 break;
             }
 
+            if (Thread.interrupted()) {
+                throw new InterruptedException();
+            }
+
             var urlMaybe = EdgeUrl.parse(doc.url);
             if (urlMaybe.isEmpty())
                 continue;
@@ -78,6 +83,7 @@ public class CrawlerRevisitor {
                 continue;
             }
 
+            size++;
 
             double skipProb;
 
@@ -150,6 +156,8 @@ public class CrawlerRevisitor {
             }
         }
 
-        return recrawled;
+        return new RecrawlMetadata(size, errors, skipped);
     }
+
+    public record RecrawlMetadata(int size, int errors, int skipped) {}
 }

code/processes/crawling-process/test/nu/marginalia/crawl/DomainStateDbTest.java

@@ -8,6 +8,7 @@ import java.io.IOException;
 import java.nio.file.Files;
 import java.nio.file.Path;
 import java.sql.SQLException;
+import java.time.Duration;
 import java.time.Instant;
 
 import static org.junit.jupiter.api.Assertions.*;
@@ -47,8 +48,8 @@ class DomainStateDbTest {
             db.save(allFields);
             db.save(minFields);
 
-            assertEquals(allFields, db.get("all.marginalia.nu").orElseThrow());
-            assertEquals(minFields, db.get("min.marginalia.nu").orElseThrow());
+            assertEquals(allFields, db.getSummary("all.marginalia.nu").orElseThrow());
+            assertEquals(minFields, db.getSummary("min.marginalia.nu").orElseThrow());
 
             var updatedAllFields = new DomainStateDb.SummaryRecord(
                     "all.marginalia.nu",
@@ -59,7 +60,19 @@ class DomainStateDbTest {
             );
 
             db.save(updatedAllFields);
-            assertEquals(updatedAllFields, db.get("all.marginalia.nu").orElseThrow());
+            assertEquals(updatedAllFields, db.getSummary("all.marginalia.nu").orElseThrow());
+        }
+    }
+
+    @Test
+    public void testMetadata() throws SQLException {
+        try (var db = new DomainStateDb(tempFile)) {
+            var original = new DomainStateDb.CrawlMeta("example.com", Instant.ofEpochMilli(12345), Duration.ofMillis(30), Duration.ofMillis(300), 1, 2, 3);
+            db.save(original);
+
+            var maybeMeta = db.getMeta("example.com");
+            assertTrue(maybeMeta.isPresent());
+            assertEquals(original, maybeMeta.get());
         }
     }