Resolve conflicts

This commit is contained in:
Serafeim Chatzopoulos 2023-05-15 15:53:12 +03:00
commit 82e2a96f51
1 changed files with 187 additions and 70 deletions

View File

@ -46,6 +46,7 @@
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>create_openaire_ranking_graph.py</jar> <jar>create_openaire_ranking_graph.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkHighExecutorMemory} --executor-memory=${sparkHighExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -56,10 +57,20 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkHighExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkHighDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<!-- The openaire graph data from which to read relations and objects --> <!-- The openaire graph data from which to read relations and objects -->
<arg>${openaireDataInput}</arg> <arg>${openaireDataInput}</arg>
<!-- Year for filtering entries w/ larger values / empty --> <!-- Year for filtering entries w/ larger values / empty -->
<arg>${currentYear}</arg> <arg>${currentYear}</arg>
<!-- number of partitions to be used on joins --> <!-- number of partitions to be used on joins -->
<arg>${sparkShufflePartitions}</arg> <arg>${sparkShufflePartitions}</arg>
@ -68,14 +79,14 @@
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/create_openaire_ranking_graph.py#create_openaire_ranking_graph.py</file> <file>${wfAppPath}/create_openaire_ranking_graph.py#create_openaire_ranking_graph.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="end" /> <ok to="end" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="openaire-graph-error" /> <error to="openaire-graph-error" />
</action> </action>
<!-- Citation Count and RAM are calculated in parallel--> <!-- Citation Count and RAM are calculated in parallel-->
<!-- Impulse Requires resources and will be run after--> <!-- Impulse Requires resources and will be run after-->
<fork name="non-iterative-rankings"> <fork name="non-iterative-rankings">
@ -83,7 +94,7 @@
<!-- <path start="spark-impulse"/> --> <!-- <path start="spark-impulse"/> -->
<path start="spark-ram"/> <path start="spark-ram"/>
</fork> </fork>
<!-- CC here --> <!-- CC here -->
<action name="spark-cc"> <action name="spark-cc">
<!-- This is required as a tag for spark jobs, regardless of programming language --> <!-- This is required as a tag for spark jobs, regardless of programming language -->
@ -96,12 +107,13 @@
<!-- using configs from an example on openaire --> <!-- using configs from an example on openaire -->
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<!-- This is the name of our job --> <!-- This is the name of our job -->
<name>Spark CC</name> <name>Spark CC</name>
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>CC.py</jar> <jar>CC.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkHighExecutorMemory} --executor-memory=${sparkHighExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -112,6 +124,16 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkHighExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkNormalDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<arg>${openaireGraphInputPath}</arg> <arg>${openaireGraphInputPath}</arg>
<!-- number of partitions to be used on joins --> <!-- number of partitions to be used on joins -->
@ -119,13 +141,13 @@
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/CC.py#CC.py</file> <file>${wfAppPath}/CC.py#CC.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="join-non-iterative-rankings" /> <ok to="join-non-iterative-rankings" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="cc-fail" /> <error to="cc-fail" />
</action> </action>
<!-- IMPULSE here --> <!-- IMPULSE here -->
<action name="spark-ram"> <action name="spark-ram">
@ -135,16 +157,17 @@
<job-tracker>${jobTracker}</job-tracker> <job-tracker>${jobTracker}</job-tracker>
<!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties --> <!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties -->
<name-node>${nameNode}</name-node> <name-node>${nameNode}</name-node>
<!-- using configs from an example on openaire --> <!-- using configs from an example on openaire -->
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<!-- This is the name of our job --> <!-- This is the name of our job -->
<name>Spark RAM</name> <name>Spark RAM</name>
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>TAR.py</jar> <jar>TAR.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkHighExecutorMemory} --executor-memory=${sparkHighExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -155,6 +178,16 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkHighExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkNormalDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<arg>${openaireGraphInputPath}</arg> <arg>${openaireGraphInputPath}</arg>
<arg>${ramGamma}</arg> <arg>${ramGamma}</arg>
@ -166,17 +199,17 @@
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/TAR.py#TAR.py</file> <file>${wfAppPath}/TAR.py#TAR.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="join-non-iterative-rankings" /> <ok to="join-non-iterative-rankings" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="ram-fail" /> <error to="ram-fail" />
</action> </action>
<!-- JOIN NON-ITERATIVE METHODS AND THEN CONTINUE TO ITERATIVE ONES --> <!-- JOIN NON-ITERATIVE METHODS AND THEN CONTINUE TO ITERATIVE ONES -->
<join name="join-non-iterative-rankings" to="spark-impulse"/> <join name="join-non-iterative-rankings" to="spark-impulse"/>
<!-- IMPULSE here --> <!-- IMPULSE here -->
<action name="spark-impulse"> <action name="spark-impulse">
<!-- This is required as a tag for spark jobs, regardless of programming language --> <!-- This is required as a tag for spark jobs, regardless of programming language -->
@ -189,12 +222,13 @@
<!-- using configs from an example on openaire --> <!-- using configs from an example on openaire -->
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<!-- This is the name of our job --> <!-- This is the name of our job -->
<name>Spark Impulse</name> <name>Spark Impulse</name>
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>CC.py</jar> <jar>CC.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkHighExecutorMemory} --executor-memory=${sparkHighExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -205,6 +239,16 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkHighExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkNormalDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<arg>${openaireGraphInputPath}</arg> <arg>${openaireGraphInputPath}</arg>
<!-- number of partitions to be used on joins --> <!-- number of partitions to be used on joins -->
@ -213,13 +257,13 @@
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/CC.py#CC.py</file> <file>${wfAppPath}/CC.py#CC.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="iterative-rankings" /> <ok to="iterative-rankings" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="impulse-fail" /> <error to="impulse-fail" />
</action> </action>
<fork name="iterative-rankings"> <fork name="iterative-rankings">
<path start="spark-pagerank"/> <path start="spark-pagerank"/>
@ -234,7 +278,7 @@
<job-tracker>${jobTracker}</job-tracker> <job-tracker>${jobTracker}</job-tracker>
<!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties --> <!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties -->
<name-node>${nameNode}</name-node> <name-node>${nameNode}</name-node>
<!-- we could add map-reduce configs here, but I don't know if we need them --> <!-- we could add map-reduce configs here, but I don't know if we need them -->
<!-- This is the type of master-client configuration for running spark --> <!-- This is the type of master-client configuration for running spark -->
<!-- <master>yarn-client</master> --> <!-- <master>yarn-client</master> -->
@ -242,16 +286,17 @@
<!-- <master>local[*]</master> --> <!-- <master>local[*]</master> -->
<!-- Reference says: The mode element if present indicates the mode of spark, where to run spark driver program. Ex: client,cluster. | In my case I always have a client --> <!-- Reference says: The mode element if present indicates the mode of spark, where to run spark driver program. Ex: client,cluster. | In my case I always have a client -->
<!-- <mode>client</mode> --> <!-- <mode>client</mode> -->
<!-- using configs from an example on openaire --> <!-- using configs from an example on openaire -->
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<!-- This is the name of our job --> <!-- This is the name of our job -->
<name>Spark Pagerank</name> <name>Spark Pagerank</name>
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>PageRank.py</jar> <jar>PageRank.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkHighExecutorMemory} --executor-memory=${sparkHighExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -262,6 +307,16 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkHighExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkNormalDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<arg>${openaireGraphInputPath}</arg> <arg>${openaireGraphInputPath}</arg>
<arg>${pageRankAlpha}</arg> <arg>${pageRankAlpha}</arg>
@ -273,14 +328,14 @@
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/PageRank.py#PageRank.py</file> <file>${wfAppPath}/PageRank.py#PageRank.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="join-iterative-rankings" /> <ok to="join-iterative-rankings" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="pagerank-fail" /> <error to="pagerank-fail" />
</action> </action>
<!-- ATTRANK here --> <!-- ATTRANK here -->
<action name="spark-attrank"> <action name="spark-attrank">
<!-- This is required as a tag for spark jobs, regardless of programming language --> <!-- This is required as a tag for spark jobs, regardless of programming language -->
@ -289,16 +344,17 @@
<job-tracker>${jobTracker}</job-tracker> <job-tracker>${jobTracker}</job-tracker>
<!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties --> <!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties -->
<name-node>${nameNode}</name-node> <name-node>${nameNode}</name-node>
<!-- using configs from an example on openaire --> <!-- using configs from an example on openaire -->
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<!-- This is the name of our job --> <!-- This is the name of our job -->
<name>Spark AttRank</name> <name>Spark AttRank</name>
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>AttRank.py</jar> <jar>AttRank.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkHighExecutorMemory} --executor-memory=${sparkHighExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -309,6 +365,16 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkHighExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkNormalDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<arg>${openaireGraphInputPath}</arg> <arg>${openaireGraphInputPath}</arg>
<arg>${attrankAlpha}</arg> <arg>${attrankAlpha}</arg>
@ -325,17 +391,17 @@
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/AttRank.py#AttRank.py</file> <file>${wfAppPath}/AttRank.py#AttRank.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="join-iterative-rankings" /> <ok to="join-iterative-rankings" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="attrank-fail" /> <error to="attrank-fail" />
</action> </action>
<!-- JOIN ITERATIVE METHODS AND THEN END --> <!-- JOIN ITERATIVE METHODS AND THEN END -->
<join name="join-iterative-rankings" to="get-file-names"/> <join name="join-iterative-rankings" to="get-file-names"/>
<!-- This will be a shell action that will output key-value pairs for output files --> <!-- This will be a shell action that will output key-value pairs for output files -->
<action name="get-file-names"> <action name="get-file-names">
@ -345,35 +411,35 @@
<job-tracker>${jobTracker}</job-tracker> <job-tracker>${jobTracker}</job-tracker>
<!-- This should give the machine/root of the hdfs --> <!-- This should give the machine/root of the hdfs -->
<name-node>${nameNode}</name-node> <name-node>${nameNode}</name-node>
<!-- Exec is needed for shell commands - points to type of shell command --> <!-- Exec is needed for shell commands - points to type of shell command -->
<exec>/usr/bin/bash</exec> <exec>/usr/bin/bash</exec>
<!-- name of script to run --> <!-- name of script to run -->
<argument>get_ranking_files.sh</argument> <argument>get_ranking_files.sh</argument>
<!-- We only pass the directory where we expect to find the rankings --> <!-- We only pass the directory where we expect to find the rankings -->
<argument>/${workingDir}</argument> <argument>/${workingDir}</argument>
<!-- the name of the file run --> <!-- the name of the file run -->
<file>${wfAppPath}/get_ranking_files.sh#get_ranking_files.sh</file> <file>${wfAppPath}/get_ranking_files.sh#get_ranking_files.sh</file>
<!-- Get the output in order to be usable by following actions --> <!-- Get the output in order to be usable by following actions -->
<capture-output/> <capture-output/>
</shell> </shell>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="format-result-files" /> <ok to="format-result-files" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="filename-getting-error" /> <error to="filename-getting-error" />
</action> </action>
<!-- Now we will run in parallel the formatting of ranking files for BiP! DB and openaire (json files) --> <!-- Now we will run in parallel the formatting of ranking files for BiP! DB and openaire (json files) -->
<fork name="format-result-files"> <fork name="format-result-files">
<path start="format-bip-files"/> <path start="format-bip-files"/>
<path start="format-json-files"/> <path start="format-json-files"/>
</fork> </fork>
<!-- Format json files --> <!-- Format json files -->
<!-- Two parts: a) format files b) make the file endings .json.gz --> <!-- Two parts: a) format files b) make the file endings .json.gz -->
<action name="format-json-files"> <action name="format-json-files">
@ -383,16 +449,17 @@
<job-tracker>${jobTracker}</job-tracker> <job-tracker>${jobTracker}</job-tracker>
<!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties --> <!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties -->
<name-node>${nameNode}</name-node> <name-node>${nameNode}</name-node>
<!-- using configs from an example on openaire --> <!-- using configs from an example on openaire -->
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<!-- This is the name of our job --> <!-- This is the name of our job -->
<name>Format Ranking Results JSON</name> <name>Format Ranking Results JSON</name>
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>format_ranking_results.py</jar> <jar>format_ranking_results.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkNormalExecutorMemory} --executor-memory=${sparkNormalExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -403,6 +470,16 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkNormalExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkNormalDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<arg>json-5-way</arg> <arg>json-5-way</arg>
<!-- Input files must be identified dynamically --> <!-- Input files must be identified dynamically -->
@ -417,14 +494,14 @@
<arg>openaire</arg> <arg>openaire</arg>
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/format_ranking_results.py#format_ranking_results.py</file> <file>${wfAppPath}/format_ranking_results.py#format_ranking_results.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="join-file-formatting" /> <ok to="join-file-formatting" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="json-formatting-fail" /> <error to="json-formatting-fail" />
</action> </action>
<!-- This is the second line of parallel workflow execution where we create the BiP! DB files --> <!-- This is the second line of parallel workflow execution where we create the BiP! DB files -->
<action name="format-bip-files"> <action name="format-bip-files">
<!-- This is required as a tag for spark jobs, regardless of programming language --> <!-- This is required as a tag for spark jobs, regardless of programming language -->
@ -433,16 +510,17 @@
<job-tracker>${jobTracker}</job-tracker> <job-tracker>${jobTracker}</job-tracker>
<!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties --> <!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties -->
<name-node>${nameNode}</name-node> <name-node>${nameNode}</name-node>
<!-- using configs from an example on openaire --> <!-- using configs from an example on openaire -->
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<!-- This is the name of our job --> <!-- This is the name of our job -->
<name>Format Ranking Results BiP! DB</name> <name>Format Ranking Results BiP! DB</name>
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>format_ranking_results.py</jar> <jar>format_ranking_results.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkNormalExecutorMemory} --executor-memory=${sparkNormalExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -453,6 +531,16 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkNormalExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkNormalDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<arg>zenodo</arg> <arg>zenodo</arg>
<!-- Input files must be identified dynamically --> <!-- Input files must be identified dynamically -->
@ -467,16 +555,16 @@
<arg>openaire</arg> <arg>openaire</arg>
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/format_ranking_results.py#format_ranking_results.py</file> <file>${wfAppPath}/format_ranking_results.py#format_ranking_results.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="join-file-formatting" /> <ok to="join-file-formatting" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="bip-formatting-fail" /> <error to="bip-formatting-fail" />
</action> </action>
<!-- Finish formatting data and end --> <!-- Finish formatting data and end -->
<join name="join-file-formatting" to="map-openaire-to-doi"/> <join name="join-file-formatting" to="map-openaire-to-doi"/>
<!-- Script here written by Serafeim: maps openaire ids to their synonyms --> <!-- Script here written by Serafeim: maps openaire ids to their synonyms -->
<action name="map-openaire-to-doi"> <action name="map-openaire-to-doi">
@ -490,16 +578,17 @@
<prepare> <prepare>
<delete path="${synonymFolder}"/> <delete path="${synonymFolder}"/>
</prepare> </prepare>
<!-- using configs from an example on openaire --> <!-- using configs from an example on openaire -->
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<!-- This is the name of our job --> <!-- This is the name of our job -->
<name>Openaire-DOI synonym collection</name> <name>Openaire-DOI synonym collection</name>
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>map_openaire_ids_to_dois.py</jar> <jar>map_openaire_ids_to_dois.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkHighExecutorMemory} --executor-memory=${sparkHighExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -510,6 +599,16 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkHighExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkHighDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<arg>${openaireDataInput}</arg> <arg>${openaireDataInput}</arg>
<!-- number of partitions to be used on joins --> <!-- number of partitions to be used on joins -->
@ -517,13 +616,13 @@
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/map_openaire_ids_to_dois.py#map_openaire_ids_to_dois.py</file> <file>${wfAppPath}/map_openaire_ids_to_dois.py#map_openaire_ids_to_dois.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="map-scores-to-dois" /> <ok to="map-scores-to-dois" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="synonym-collection-fail" /> <error to="synonym-collection-fail" />
</action> </action>
<!-- Script here written by Serafeim: maps openaire ids to their synonyms --> <!-- Script here written by Serafeim: maps openaire ids to their synonyms -->
@ -535,15 +634,16 @@
<!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties --> <!-- This should give the machine/root of the hdfs, serafeim has provided a link with the required job properties -->
<name-node>${nameNode}</name-node> <name-node>${nameNode}</name-node>
<!-- using configs from an example on openaire --> <!-- using configs from an example on openaire -->
<master>yarn-cluster</master> <master>yarn-cluster</master>
<mode>cluster</mode> <mode>cluster</mode>
<!-- This is the name of our job --> <!-- This is the name of our job -->
<name>Mapping Openaire Scores to DOIs</name> <name>Mapping Openaire Scores to DOIs</name>
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>map_scores_to_dois.py</jar> <jar>map_scores_to_dois.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkHighExecutorMemory} --executor-memory=${sparkHighExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
@ -554,6 +654,16 @@
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir} --conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}
</spark-opts> </spark-opts>
=======
<spark-opts>--executor-memory ${sparkHighExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkHighDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
--conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}
--conf spark.eventLog.dir=${nameNode}${spark2EventLogDir}</spark-opts>
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
<!-- Script arguments here --> <!-- Script arguments here -->
<arg>${synonymFolder}</arg> <arg>${synonymFolder}</arg>
<!-- Number of partitions --> <!-- Number of partitions -->
@ -568,13 +678,13 @@
<!-- This needs to point to the file on the hdfs i think --> <!-- This needs to point to the file on the hdfs i think -->
<file>${wfAppPath}/map_scores_to_dois.py#map_scores_to_dois.py</file> <file>${wfAppPath}/map_scores_to_dois.py#map_scores_to_dois.py</file>
</spark> </spark>
<!-- Do this after finishing okay --> <!-- Do this after finishing okay -->
<ok to="deleteOutputPathForActionSet" /> <ok to="deleteOutputPathForActionSet" />
<!-- Go there if we have an error --> <!-- Go there if we have an error -->
<error to="map-scores-fail" /> <error to="map-scores-fail" />
</action> </action>
<action name="deleteOutputPathForActionSet"> <action name="deleteOutputPathForActionSet">
<fs> <fs>
@ -629,11 +739,18 @@
<!-- Script name goes here --> <!-- Script name goes here -->
<jar>projects_impact.py</jar> <jar>projects_impact.py</jar>
<!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro --> <!-- spark configuration options: I've taken most of them from an example from dhp workflows / Master value stolen from sandro -->
<<<<<<< HEAD
<spark-opts> <spark-opts>
--executor-memory=${sparkHighExecutorMemory} --executor-memory=${sparkHighExecutorMemory}
--executor-cores=${sparkExecutorCores} --executor-cores=${sparkExecutorCores}
--driver-memory=${sparkNormalDriverMemory} --driver-memory=${sparkNormalDriverMemory}
--conf spark.sql.shuffle.partitions=${sparkShufflePartitions} --conf spark.sql.shuffle.partitions=${sparkShufflePartitions}
=======
<spark-opts>--executor-memory ${sparkHighExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory ${sparkNormalDriverMemory}
--master yarn
--deploy-mode cluster
--conf spark.sql.shuffle.partitions=7680
>>>>>>> 4a905932a3db36c61570c24b9aa54283cd30abba
--conf spark.extraListeners=${spark2ExtraListeners} --conf spark.extraListeners=${spark2ExtraListeners}
--conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners} --conf spark.sql.queryExecutionListeners=${spark2SqlQueryExecutionListeners}
--conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress} --conf spark.yarn.historyServer.address=${spark2YarnHistoryServerAddress}