michal.politowski
/
UrlsController
forked from lsmyrnaios/UrlsController
1
0
Fork 0
Code Pull Requests Activity

- Create the HDFS-subDirs before generating "callableTasks" for creating and uploading the parquetFiles. 

- Delete gradle .zip file after installation.
This commit is contained in:
Lampros Smyrnaios 2023-08-04 15:30:41 +03:00
parent 860c73ea91
commit a524375656
3 changed files with 5 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
installAndRun.sh
View File

@ -35,6 +35,7 @@ if [[ justInstall -eq 0 ]]; then
echo -e "\nAsking for sudo, in order to install 'gradle'..\n"
sudo mkdir /opt/gradle
sudo apt install -y unzip && sudo unzip -d /opt/gradle gradle-${gradleVersion}-bin.zip
sudo rm -rf gradle-${gradleVersion}-bin.zip
#ls /opt/gradle/gradle-${gradleVersion} # For debugging installation
fi

2
src/main/java/eu/openaire/urls_controller/controllers/BulkImportController.java
View File

@ -201,7 +201,7 @@ public class BulkImportController {
bulkImportService.bulkImportFullTextsFromDirectory(bulkImportReport, finalRelativeBulkImportDir, finalBulkImportDir, givenDir, provenance, bulkImportSource, shouldDeleteFilesOnFinish)
);
return ResponseEntity.ok().body(new BulkImportResponse(msg, bulkImportReportID));
return ResponseEntity.ok().body(new BulkImportResponse(msg, bulkImportReportID)); // The response is automatically serialized to json and it's of type "application/json".
}

6
src/main/java/eu/openaire/urls_controller/services/UrlsServiceImpl.java
View File

@ -311,17 +311,17 @@ public class UrlsServiceImpl implements UrlsService {
logger.debug("Going to write the results in the parquet files, then upload them to HDFS, and then load them into the database's tables. For batch-assignments_" + curReportAssignmentsCounter);
List<Callable<ParquetReport>> callableTasks = parquetFileUtils.getTasksForCreatingAndUploadingParquetFiles(urlReports, sizeOfUrlReports, curReportAssignmentsCounter, localParquetPath, uploadFullTextsResponse);
// Create HDFS subDirs for these assignments. Other background threads handling other assignments will not interfere with loading of parquetFiles to the DB tables.
String endingMkDirAndParams = curReportAssignmentsCounter + "/" + parquetFileUtils.mkDirsAndParams;
if ( !parquetFileUtils.applyHDFOperation(parquetFileUtils.webHDFSBaseUrl + parquetFileUtils.parquetHDFSDirectoryPathAttempts + endingMkDirAndParams)
|| !parquetFileUtils.applyHDFOperation(parquetFileUtils.webHDFSBaseUrl + parquetFileUtils.parquetHDFSDirectoryPathPayloadsAggregated + endingMkDirAndParams) )
|| !parquetFileUtils.applyHDFOperation(parquetFileUtils.webHDFSBaseUrl + parquetFileUtils.parquetHDFSDirectoryPathPayloadsAggregated + endingMkDirAndParams) )
{
postReportResultToWorker(curWorkerId, curReportAssignmentsCounter, "Error when creating the HDFS sub-directories for assignments_" + curReportAssignmentsCounter);
return false;
}
List<Callable<ParquetReport>> callableTasks = parquetFileUtils.getTasksForCreatingAndUploadingParquetFiles(urlReports, sizeOfUrlReports, curReportAssignmentsCounter, localParquetPath, uploadFullTextsResponse);
boolean hasAttemptParquetFileProblem = false;
boolean hasPayloadParquetFileProblem = false;