From a9b1b20a51287bd59387018fcb08c29ad351b2b5 Mon Sep 17 00:00:00 2001
From: LSmyrnaios <lsmyrnaios@gmail.com>
Date: Tue, 30 May 2023 17:58:29 +0300
Subject: [PATCH] - Prevent running out of space, by checking the available
 free space and stalling the acquisition of new assignments until more free
 space becomes available. - Fix missing change.

---
 installAndRun.sh                              |  2 +-
 .../components/ScheduledTasks.java            | 31 +++++++++++++++++++
 2 files changed, 32 insertions(+), 1 deletion(-)

diff --git a/installAndRun.sh b/installAndRun.sh
index c0e0e1d..1359d5f 100755
--- a/installAndRun.sh
+++ b/installAndRun.sh
@@ -29,7 +29,7 @@ shouldBeCarefulWithMaxHeap=0 # This is NOT a cmd-arg.
 if [[ justRun -eq 0 ]]; then
 
   if [[ avoidReInstallingPublicationsRetriever -eq 1 ]]; then
-    if [[ ! -f ./libs/publications_retriever-1.0-SNAPSHOT.jar ]]; then
+    if [[ ! -f ./libs/publications_retriever-1.1-SNAPSHOT.jar ]]; then
       echo -e "\n\nThe required \"PublicationsRetriever\" software has not been installed yet, thus the script will override the user-defined value of \"avoidReInstallingPublicationsRetriever\" to FALSE..\n\n"
       avoidReInstallingPublicationsRetriever=0; # In case the jar-file does not exists, then make sure we follow the normal procedure, independently from what the user requested.
     fi
diff --git a/src/main/java/eu/openaire/urls_worker/components/ScheduledTasks.java b/src/main/java/eu/openaire/urls_worker/components/ScheduledTasks.java
index 8414590..bef39fa 100644
--- a/src/main/java/eu/openaire/urls_worker/components/ScheduledTasks.java
+++ b/src/main/java/eu/openaire/urls_worker/components/ScheduledTasks.java
@@ -38,6 +38,20 @@ public class ScheduledTasks {
     @Value("${workerReportsDirPath}")
     private String workerReportsDirPath;
 
+    private static final File rootPath = new File("/");
+    private static final long oneAndHalfGB = 1_610_612_736L;    // 1.5 GB free space per 1.000-assignments-batch.
+    private static long requiredFreeSpace;
+
+
+    public ScheduledTasks(@Value("${info.maxAssignmentsLimitPerBatch}") int maxAssignmentsLimitPerBatch)
+    {
+        if ( maxAssignmentsLimitPerBatch < 1_000 )
+            requiredFreeSpace = oneAndHalfGB;
+        else
+            requiredFreeSpace = oneAndHalfGB * (maxAssignmentsLimitPerBatch / 1_000);
+
+        logger.info("The \"requiredFreeSpace\" for the app to request new assignments, having \"maxAssignmentsLimitPerBatch\" equal to " + maxAssignmentsLimitPerBatch + " , is: " + (requiredFreeSpace / (1024 * 1024)) + " Mb");
+    }
 
     @Scheduled(fixedDelay = 1)  // Request the next batch immediately after the last one finishes.
     public void handleNewAssignments() {
@@ -47,6 +61,23 @@ public class ScheduledTasks {
             return;
         }
 
+        if ( rootPath.getFreeSpace() < requiredFreeSpace ) {
+            // It's not safe to proceed with downloading more files and risk of "noSpaceLeft" error.
+            // Wait for the Controller to take the full-texts and any remaining files to be deleted, so that more free-space becomes available.
+            // We need to have some buffer zone for the ".tar" files which will be created from the already downloaded full-texts, when the Controller starts requesting them.
+            if ( GeneralController.shouldShutdownWorker ) { // Make sure the worker shuts-down, in case the user sends the relevant request, while the worker is stuck in a free-space check loop.
+                AssignmentsHandler.shouldNotRequestMore = true;
+                return;
+            }
+            logger.warn("The free space is running out (less than " + (requiredFreeSpace / (1024 * 1024)) + " Mb). Will avoid to get new assignments for the next 15 minutes.");
+            try {
+                Thread.sleep(900_000);  // Sleep for 15 mins to stall the scheduler from retrying right away, thus giving time to the disk-space to be freed.
+            } catch (InterruptedException ie) {
+                logger.warn("Sleeping was interrupted!");
+            }
+            return; // Cause this method to be called again, so that the Free-space can be checked again before proceeding with new assignments.
+        }
+
         if ( AssignmentsHandler.hadConnectionErrorOnRequest ) {
             if ( GeneralController.shouldShutdownWorker ) { // Make sure the worker shuts-down, in case the user sends the relevant request, while the worker is stuck in a data-request error-loop.
                 AssignmentsHandler.shouldNotRequestMore = true;