Prevent new gathers if working area is low on free space

This should make crawls queue up rather than creating a whole of failed
(or worse, 'successful' but empty) crawls in the case the working area
runs out of space.

gatherer/src/pandas/gatherer/core/Config.java

@@ -40,6 +40,9 @@ public class Config {
     private String oidcClientSecret;
     private Path legacyScripts;
 
+    // Browsertrix stops at 10%, so let's ensure there's a little more than that
+    private double minFreeSpacePercent = 11.0;
+
     public Config() {
     }
 
@@ -284,4 +287,12 @@ public Path getLegacyScripts() {
     public void setLegacyScripts(Path legacyScripts) {
         this.legacyScripts = legacyScripts;
     }
+
+    public double getMinFreeSpacePercent() {
+        return minFreeSpacePercent;
+    }
+
+    public void setMinFreeSpacePercent(double minFreeSpacePercent) {
+        this.minFreeSpacePercent = minFreeSpacePercent;
+    }
 }

gatherer/src/pandas/gatherer/core/GatherManager.java

@@ -3,7 +3,6 @@
 import jakarta.annotation.PreDestroy;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
-import org.springframework.context.Lifecycle;
 import org.springframework.context.SmartLifecycle;
 import org.springframework.stereotype.Component;
 import pandas.collection.Title;
@@ -44,6 +43,7 @@ public class GatherManager implements AutoCloseable, SmartLifecycle {
 	private volatile boolean paused;
 	private boolean running;
 	private ScheduledFuture<?> updateGatherStatsTask;
+	private long lastSpaceWarningTime = 0;
 
 	public GatherManager(Config config, WorkingArea workingArea, InstanceRepository instanceRepository,
 						 TitleRepository titleRepository, InstanceService instanceService,
@@ -123,6 +123,19 @@ Instance nextInstance(String gatherMethod, String threadName) {
 				return instance;
 			}
 
+			// Ensure there's enough free space in working area to allow starting of new gathers.
+			double freeSpacePercent = workingArea.getFreeSpacePercent();
+			if (freeSpacePercent < config.getMinFreeSpacePercent()) {
+				long now = System.currentTimeMillis();
+				if (lastSpaceWarningTime + 1000 * 60 < now) {
+					lastSpaceWarningTime = now;
+					log.warn("Preventing new gathers because free space percentage for {} is below the minimum " +
+							 "threshold ({}%). Current: {}%",
+							workingArea.getPath(), config.getMinFreeSpacePercent(), freeSpacePercent);
+				}
+				return null;
+			}
+
 			// now look for titles scheduled for a new gather
 			// XXX: we ignore any titles that were last gathered within the current minute
 			//      this is to ensure that we don't generate an instance with the same datestring

gatherer/src/pandas/gatherer/core/WorkingArea.java

@@ -7,6 +7,7 @@
 import pandas.gatherer.httrack.Pandora2Warc;
 import pandas.gatherer.repository.Repository;
 
+import java.io.File;
 import java.io.IOException;
 import java.nio.file.*;
 import java.nio.file.attribute.BasicFileAttributes;
@@ -264,4 +265,21 @@ public void cleanupToDelete() {
     public Path getInstanceDir(long pi, String dateString) {
         return workingdir.resolve(String.valueOf(pi)).resolve(dateString).toAbsolutePath();
     }
+
+    public double getFreeSpacePercent() {
+        File file = workingdir.toFile();
+        long totalSpace = file.getTotalSpace();
+        long freeSpace = file.getFreeSpace();
+
+        if (totalSpace == 0) {
+            log.warn("Unable to calculate free space percentage. Total space is 0.");
+            return 50.0;
+        }
+
+        return ((double) freeSpace / totalSpace) * 100;
+    }
+
+    public Path getPath() {
+        return workingdir;
+    }
 }