spyros
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Correct parameter parsing

This commit is contained in:
Spyros Zoupanos 2020-10-01 23:24:40 +03:00
parent dc6114a24e
commit c1257ac7c5
10 changed files with 235 additions and 62 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
dhp-workflows/dhp-usage-stats-update/job-override.properties.bk Normal file
View File

@ -0,0 +1,7 @@
stats_db_name=openaire_beta_20200618_stats
openaire_db_name=openaire_beta_20200618
external_stats_db_name=stats_ext
stats_db_shadow_name=openaire_beta_20200618_stats_shadow
hive_timeout=3000
hive_spark_client_timeout=100000
hive_spark_client_server_timeout=100000

25
dhp-workflows/dhp-usage-stats-update/pom.xml
View File

@ -2,16 +2,28 @@
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
<parent>
<!-- <parent>
<artifactId>dhp-workflows</artifactId >
<groupId>eu.dnetlib.dhp</groupId>
<version>1.1.7-SNAPSHOT</version>
</parent>
<groupId>eu.dnetlib</groupId>
<groupId>eu.dnetlib</groupId> -->
<!-- <parent>
<groupId>eu.dnetlib.dhp</groupId>
<artifactId>dhp-workflows</artifactId>
<version>1.1.7-SNAPSHOT</version>
</parent>
<modelVersion>4.0.0</modelVersion>
<artifactId>dhp-usage-stats-update</artifactId> -->
<parent>
<artifactId>dhp-workflows</artifactId>
<groupId>eu.dnetlib.dhp</groupId>
<version>1.1.7-SNAPSHOT</version>
</parent>
<modelVersion>4.0.0</modelVersion>
<artifactId>dhp-usage-stats-update</artifactId>
<properties>
<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
@ -45,13 +57,16 @@
<groupId>org.apache.hive</groupId>
<artifactId>hive-jdbc</artifactId>
<version>${cdh.hive.version}</version>
<!-- <version>3.1.2</version> -->
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-common</artifactId>
<version>${cdh.hadoop.version}</version>
</dependency>
<dependency>
<groupId>eu.dnetlib.dhp</groupId>
<artifactId>dhp-common</artifactId>
<version>${project.version}</version>
</dependency>
</dependencies>
</project>

53
dhp-workflows/dhp-usage-stats-update/src/main/java/eu/dnetlib/oa/graph/usagestats/export/ExecuteWorkflow.java
View File

@ -9,20 +9,59 @@ package eu.dnetlib.oa.graph.usagestats.export;
import java.io.InputStream;
import java.util.Properties;
import org.apache.commons.io.IOUtils;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
/**
* @author dpie
* @author dpie, Spyros Zoupanos
*/
public class ExecuteWorkflow {
static String matomoAuthToken;
static String matomoBaseURL;
static String repoLogPath;
static String portalLogPath;
static String portalMatomoID;
static String irusUKBaseURL;
static String irusUKReportPath;
static String sarcsReportPathArray;
static String sarcsReportPathNonArray;
static String lareferenciaLogPath;
static String lareferenciaBaseURL;
static String lareferenciaAuthToken;
public static void main(String args[]) throws Exception {
Properties prop = new Properties();
InputStream propertiesInputStream = UsageStatsExporter.class
.getClassLoader()
.getResourceAsStream("usagestats.properties");
prop.load(propertiesInputStream);
// Properties prop = new Properties();
// InputStream propertiesInputStream = UsageStatsExporter.class
// .getClassLoader()
// .getResourceAsStream("usagestats.properties");
// prop.load(propertiesInputStream);
UsageStatsExporter usagestatsExport = new UsageStatsExporter(prop);
final ArgumentApplicationParser parser = new ArgumentApplicationParser(
IOUtils
.toString(
UsageStatsExporter.class
.getResourceAsStream(
// "/eu/dnetlib/dhp/actionmanager/migration/migrate_actionsets_parameters.json")));
"/eu/dnetlib/dhp/oa/graph/usagestats/export/usagestats_parameters.json")));
parser.parseArgument(args);
matomoAuthToken = parser.get("matomoAuthToken");
matomoBaseURL = parser.get("matomoBaseURL");
repoLogPath = parser.get("repoLogPath");
portalLogPath = parser.get("portalLogPath");
portalMatomoID = parser.get("portalMatomoID");
irusUKBaseURL = parser.get("irusUKBaseURL");
irusUKReportPath = parser.get("irusUKReportPath");
sarcsReportPathArray = parser.get("sarcsReportPathArray");
sarcsReportPathNonArray = parser.get("sarcsReportPathNonArray");
lareferenciaLogPath = parser.get("lareferenciaLogPath");
lareferenciaBaseURL = parser.get("lareferenciaBaseURL");
lareferenciaAuthToken = parser.get("lareferenciaAuthToken");
UsageStatsExporter usagestatsExport = new UsageStatsExporter(null);
usagestatsExport.export();
}
}

2
dhp-workflows/dhp-usage-stats-update/src/main/java/eu/dnetlib/oa/graph/usagestats/export/IrusStats.java
View File

@ -149,7 +149,7 @@ public class IrusStats {
" >\n" +
")\n" +
"ROW FORMAT SERDE 'org.apache.hive.hcatalog.data.JsonSerDe'\n" +
"LOCATION '" + UsageStatsExporter.irusUKReportPath + "'\n" +
"LOCATION '" + ExecuteWorkflow.irusUKReportPath + "'\n" +
"TBLPROPERTIES (\"transactional\"=\"false\")";
stmt.executeUpdate(create_sushilogtmp_json);
System.out.println("====> Created sushilogtmp_json table");

2
dhp-workflows/dhp-usage-stats-update/src/main/java/eu/dnetlib/oa/graph/usagestats/export/LaReferenciaStats.java
View File

@ -174,7 +174,7 @@ public class LaReferenciaStats {
" >" +
")\n" +
"ROW FORMAT SERDE 'org.apache.hive.hcatalog.data.JsonSerDe'\n" +
"LOCATION '" + UsageStatsExporter.lareferenciaLogPath + "'\n" +
"LOCATION '" + ExecuteWorkflow.lareferenciaLogPath + "'\n" +
"TBLPROPERTIES (\"transactional\"=\"false\")";
stmt.executeUpdate(create_lareferencialogtmp_json);
System.out.println("====> Created lareferencialogtmp_json");

4
dhp-workflows/dhp-usage-stats-update/src/main/java/eu/dnetlib/oa/graph/usagestats/export/PiwikStatsDB.java
View File

@ -262,7 +262,7 @@ public class PiwikStatsDB {
" >\n" +
")\n" +
"ROW FORMAT SERDE 'org.apache.hive.hcatalog.data.JsonSerDe'\n" +
"LOCATION '" + UsageStatsExporter.repoLogPath + "'\n" +
"LOCATION '" + ExecuteWorkflow.repoLogPath + "'\n" +
"TBLPROPERTIES (\"transactional\"=\"false\")";
stmt.executeUpdate(create_piwiklogtmp_json);
System.out.println("====> Created piwiklogtmp_json");
@ -749,7 +749,7 @@ public class PiwikStatsDB {
" >\n" +
")\n" +
"ROW FORMAT SERDE 'org.apache.hive.hcatalog.data.JsonSerDe'\n" +
"LOCATION '" + UsageStatsExporter.repoLogPath + "'\n" +
"LOCATION '" + ExecuteWorkflow.repoLogPath + "'\n" +
"TBLPROPERTIES (\"transactional\"=\"false\")";
stmt.executeUpdate(create_process_portal_log_tmp_json);
System.out.println("====> Created process_portal_log_tmp_json");

87
dhp-workflows/dhp-usage-stats-update/src/main/java/eu/dnetlib/oa/graph/usagestats/export/UsageStatsExporter.java
View File

@ -6,27 +6,29 @@ import java.sql.SQLException;
import java.sql.Statement;
import java.util.Properties;
import org.apache.log4j.Logger;
import org.apache.commons.io.IOUtils;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
public class UsageStatsExporter {
private Logger log = Logger.getLogger(this.getClass());
private static final Logger logger = LoggerFactory.getLogger(UsageStatsExporter.class);
private Properties properties;
static String matomoAuthToken = "703bd17d845acdaf795e01bb1e0895b9";
static String matomoBaseURL = "analytics.openaire.eu";
static String repoLogPath = "/user/spyros/logs/usage_stats_logs6/Repologs";
static String portalLogPath = "/user/spyros/logs/usage_stats_logs6/Portallogs/";
static String portalMatomoID = "109";
static String irusUKBaseURL = "https://irus.jisc.ac.uk/api/sushilite/v1_7/";
static String irusUKReportPath = "/user/spyros/logs/usage_stats_logs6/irusUKReports";
static String sarcsReportPathArray = "/user/spyros/logs/usage_stats_logs6/sarcReports/Array";
static String sarcsReportPathNonArray = "/user/spyros/logs/usage_stats_logs6/sarcReports/NonArray";
static String lareferenciaLogPath = "/user/spyros/logs/usage_stats_logs6/lareferencia";
static String lareferenciaBaseURL = "http://matomo.lareferencia.info";
static String lareferenciaAuthToken = "484874b3655d5a831eb8db33695790c4";
// static String matomoAuthToken = "703bd17d845acdaf795e01bb1e0895b9";
// static String matomoBaseURL = "analytics.openaire.eu";
// static String repoLogPath = "/user/spyros/logs/usage_stats_logs6/Repologs";
// static String portalLogPath = "/user/spyros/logs/usage_stats_logs6/Portallogs/";
// static String portalMatomoID = "109";
// static String irusUKBaseURL = "https://irus.jisc.ac.uk/api/sushilite/v1_7/";
//
// static String irusUKReportPath = "/user/spyros/logs/usage_stats_logs6/irusUKReports";
// static String sarcsReportPathArray = "/user/spyros/logs/usage_stats_logs6/sarcReports/Array";
// static String sarcsReportPathNonArray = "/user/spyros/logs/usage_stats_logs6/sarcReports/NonArray";
//
// static String lareferenciaLogPath = "/user/spyros/logs/usage_stats_logs6/lareferencia";
// static String lareferenciaBaseURL = "http://matomo.lareferencia.info";
// static String lareferenciaAuthToken = "484874b3655d5a831eb8db33695790c4";
public UsageStatsExporter(Properties properties) {
this.properties = properties;
@ -57,9 +59,51 @@ public class UsageStatsExporter {
stmt.close();
}
// public static void main(String[] args) throws Exception {
// final ArgumentApplicationParser parser = new ArgumentApplicationParser(
// IOUtils
// .toString(
// UsageStatsExporter.class
// .getResourceAsStream(
// "/eu/dnetlib/dhp/actionmanager/migration/migrate_actionsets_parameters.json")));
// parser.parseArgument(args);
//
// new UsageStatsExporter(null).run(parser);
// }
// private void run(ArgumentApplicationParser parser) throws Exception {
//
// final String isLookupUrl = parser.get("isLookupUrl");
// final String sourceNN = parser.get("sourceNameNode");
// final String targetNN = parser.get("targetNameNode");
// final String workDir = parser.get("workingDirectory");
// final Integer distcp_num_maps = Integer.parseInt(parser.get("distcp_num_maps"));
//
// final String distcp_memory_mb = parser.get("distcp_memory_mb");
// final String distcp_task_timeout = parser.get("distcp_task_timeout");
//
// final String transform_only_s = parser.get("transform_only");
// }
// public void export() throws Exception {
public void export() throws Exception {
System.out.println(ExecuteWorkflow.matomoAuthToken);
System.out.println(ExecuteWorkflow.matomoBaseURL);
System.out.println(ExecuteWorkflow.repoLogPath);
System.out.println(ExecuteWorkflow.portalLogPath);
System.out.println(ExecuteWorkflow.irusUKBaseURL);
System.out.println(ExecuteWorkflow.irusUKReportPath);
System.out.println(ExecuteWorkflow.sarcsReportPathArray);
System.out.println(ExecuteWorkflow.sarcsReportPathNonArray);
System.out.println(ExecuteWorkflow.lareferenciaLogPath);
System.out.println(ExecuteWorkflow.lareferenciaBaseURL);
System.out.println(ExecuteWorkflow.lareferenciaAuthToken);
logger.info("=====> Test of the logger (info)");
logger.debug("=====> Test of the logger (debug)");
logger.error("=====> Test of the logger (error)");
// connect to DB
System.out.println("====> Initialising DB properties");
ConnectDB.init(properties);
@ -69,12 +113,12 @@ public class UsageStatsExporter {
// Create DB tables - they are also needed to download the statistics too
System.out.println("====> Creating database and tables");
PiwikStatsDB piwikstatsdb = new PiwikStatsDB(repoLogPath, portalLogPath);
PiwikStatsDB piwikstatsdb = new PiwikStatsDB(ExecuteWorkflow.repoLogPath, ExecuteWorkflow.portalLogPath);
//
// // Download the statistics - The following 2 lines are not needed after the download - Commenting them out for
// // the moment
System.out.println("====> Initializing the download logs module");
PiwikDownloadLogs piwd = new PiwikDownloadLogs(matomoBaseURL, matomoAuthToken);
PiwikDownloadLogs piwd = new PiwikDownloadLogs(ExecuteWorkflow.matomoBaseURL, ExecuteWorkflow.matomoAuthToken);
System.out.println("====> Downloading piwik logs");
// piwd.GetOpenAIRELogs(repoLogPath, portalLogPath, portalMatomoID);
System.out.println("====> Downloaded piwik logs");
@ -85,14 +129,15 @@ public class UsageStatsExporter {
piwikstatsdb.setCounterRobotsURL(cRobotsUrl);
System.out.println("====> Processing logs");
piwikstatsdb.processLogs();
log.info("process logs done");
// log.info("process logs done");
System.out.println("====> Creating LaReferencia tables");
LaReferenciaDownloadLogs lrf = new LaReferenciaDownloadLogs(lareferenciaBaseURL, lareferenciaAuthToken);
LaReferenciaDownloadLogs lrf = new LaReferenciaDownloadLogs(ExecuteWorkflow.lareferenciaBaseURL,
ExecuteWorkflow.lareferenciaAuthToken);
System.out.println("====> Downloading LaReferencia logs");
// lrf.GetLaReferenciaRepos(lareferenciaLogPath);
System.out.println("====> Downloaded LaReferencia logs");
LaReferenciaStats lastats = new LaReferenciaStats(lareferenciaLogPath);
LaReferenciaStats lastats = new LaReferenciaStats(ExecuteWorkflow.lareferenciaLogPath);
System.out.println("====> Processing LaReferencia logs");
// lastats.processLogs();
// log.info("LaReferencia logs done");

78
dhp-workflows/dhp-usage-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/usagestats/export/usagestats_parameters.json Normal file
View File

@ -0,0 +1,78 @@
[
{
"paramName": "mat",
"paramLongName": "matomoAuthToken",
"paramDescription": "when true will stop SparkSession after job execution",
"paramRequired": false
},
{
"paramName": "mbu",
"paramLongName": "matomoBaseURL",
"paramDescription": "URL of the isLookUp Service",
"paramRequired": true
},
{
"paramName": "rlp",
"paramLongName": "repoLogPath",
"paramDescription": "nameNode of the source cluster",
"paramRequired": true
},
{
"paramName": "plp",
"paramLongName": "portalLogPath",
"paramDescription": "namoNode of the target cluster",
"paramRequired": true
},
{
"paramName": "pmi",
"paramLongName": "portalMatomoID",
"paramDescription": "namoNode of the target cluster",
"paramRequired": true
},
{
"paramName": "iukbuw",
"paramLongName": "irusUKBaseURL",
"paramDescription": "working directory",
"paramRequired": true
},
{
"paramName": "iukrp",
"paramLongName": "irusUKReportPath",
"paramDescription": "maximum number of map tasks used in the distcp process",
"paramRequired": true
},
{
"paramName": "srpa",
"paramLongName": "sarcsReportPathArray",
"paramDescription": "memory for distcp action copying actionsets from remote cluster",
"paramRequired": true
},
{
"paramName": "srpna",
"paramLongName": "sarcsReportPathNonArray",
"paramDescription": "timeout for distcp copying actions from remote cluster",
"paramRequired": true
},
{
"paramName": "llp",
"paramLongName": "lareferenciaLogPath",
"paramDescription": "activate tranform-only mode. Only apply transformation step",
"paramRequired": true
},
{
"paramName": "lbu",
"paramLongName": "lareferenciaBaseURL",
"paramDescription": "activate tranform-only mode. Only apply transformation step",
"paramRequired": true
},
{
"paramName": "lat",
"paramLongName": "lareferenciaAuthToken",
"paramDescription": "activate tranform-only mode. Only apply transformation step",
"paramRequired": true
}
]

13
dhp-workflows/dhp-usage-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/usagestats/oozie_app/workflow.xml
View File

@ -38,6 +38,19 @@
<action name='Step1'>
<java>
<main-class>eu.dnetlib.oa.graph.usagestats.export.ExecuteWorkflow</main-class>
<arg>--matomoAuthToken</arg><arg>${matomoAuthToken}</arg>
<arg>--matomoBaseURL</arg><arg>${matomoBaseURL}</arg>
<arg>--repoLogPath</arg><arg>${repoLogPath}</arg>
<arg>--portalLogPath</arg><arg>${portalLogPath}</arg>
<arg>--portalMatomoID</arg><arg>${portalMatomoID}</arg>
<arg>--irusUKBaseURL</arg><arg>${irusUKBaseURL}</arg>
<arg>--irusUKReportPath</arg><arg>${irusUKReportPath}</arg>
<arg>--sarcsReportPathArray</arg><arg>${sarcsReportPathArray}</arg>
<arg>--sarcsReportPathNonArray</arg><arg>${sarcsReportPathNonArray}</arg>
<arg>--lareferenciaLogPath</arg><arg>${lareferenciaLogPath}</arg>
<arg>--lareferenciaBaseURL</arg><arg>${lareferenciaBaseURL}</arg>
<arg>--lareferenciaAuthToken</arg><arg>${lareferenciaAuthToken}</arg>
<capture-output/>
</java>
<ok to="End" />
<error to="Kill" />

24
dhp-workflows/dhp-usage-stats-update/src/main/resources/usagestats.properties
View File

@ -1,24 +0,0 @@
#logging.config=log4j.properties
db.driverClassName=org.postgresql.Driver
#spring.datasource.url=jdbc:postgresql://88.197.53.70:5432/stats
#spring.datasource.url=jdbc:postgresql://statsdb-beta.openaire.eu:5432/stats
db.Url=jdbc:postgresql://88.197.53.70:5432/stats
#db.Url=jdbc:postgresql://statsdb-prod.openaire.eu/stats
db.username=sqoop
db.password=sqoop
db.defaultschema=usagestats
matomo.AuthToken=703bd17d845acdaf795e01bb1e0895b9
matomo.BaseUrl=analytics.openaire.eu
#repo.LogPath=/Users/dpie/Desktop/Repologs/
repo.LogPath=/user/spyros/logs/usage_stats_logs/Repologs
portal.LogPath=/user/spyros/logs/usage_stats_logs/Portallogs/
portal.MatomoID=109
COUNTER.robots.Url=https://raw.githubusercontent.com/atmire/COUNTER-Robots/master/COUNTER_Robots_list.json
IRUS_UK.BaseUrl=https://irus.jisc.ac.uk/api/sushilite/v1_7/
#server.compression.enabled=true
#compression.max_number_of_records=1000
#usagestats.redis.hostname=localhost
#usagestats.redis.port=6379
#spring.jackson.serialization.INDENT_OUTPUT=true
#download.folder=/Users/dpie/DownloadSushiLite
#sushi-lite.server=http://localhost:8080