set dynamic memory limit in pdfalto_server

kermitt2 · Aug 6, 2023 · 3de1ca4 · 3de1ca4
1 parent bbf1e85
commit 3de1ca4
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 2 deletions.
diff --git a/grobid-core/src/main/java/org/grobid/core/document/DocumentSource.java b/grobid-core/src/main/java/org/grobid/core/document/DocumentSource.java
@@ -146,6 +146,8 @@ public File pdfalto(Integer timeout, boolean force, int startPage,
             if (GrobidProperties.isContextExecutionServer()) {
                 cmd.add("--timeout");
                 cmd.add(String.valueOf(GrobidProperties.getPdfaltoTimeoutS()));
+                cmd.add("--ulimit");
+                cmd.add(String.valueOf(GrobidProperties.getPdfaltoMemoryLimitMb() * 1024));
                 tmpPathXML = processPdfaltoServerMode(pdfPath, tmpPathXML, cmd);
             } else {
                 if (!SystemUtils.IS_OS_WINDOWS && !SystemUtils.IS_OS_MAC) {

diff --git a/grobid-home/pdfalto/lin-64/pdfalto_server b/grobid-home/pdfalto/lin-64/pdfalto_server
@@ -6,26 +6,36 @@ timeout=20   # 20 seconds
 interval=1
 # Delay between posting the SIGTERM signal and destroying the process by SIGKILL.
 delay=0
+# Limit the memory usage by this bash script
+memory_limit=6242304
 
 command=${0:0:${#0}-7}
 args=("$@")
 pdfalto_params=()
 
 for ((n=0; n<="$#";n++)); do
-   case ${args[n]} in
+    case ${args[n]} in
          --timeout) 
 	      timeout=${args[n+1]}
               ((n++))
               ;;
+         --ulimit) 
+          memory_limit=${args[n+1]}
+              ((n++))
+              ;;
          *)
               pdfalto_params+=" ${args[n]}"
               ;;
-   esac
+    esac
 done
 
 #echo timeout $timeout
 #echo pdfalto commands: $command $pdfalto_params
 
+#echo memory_limit $memory_limit
+# Limit the memory usage by this bash script
+ulimit -Sv $memory_limit
+
 # kill -0 pid   Exit code indicates if a signal may be sent to $pid process.
 (
     ((t = timeout))