Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Bug rmi timeout #251

Merged
merged 4 commits into from
May 19, 2020
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions Exareme-Docker/files/root/exareme/exareme-admin.sh
Original file line number Diff line number Diff line change
Expand Up @@ -95,6 +95,12 @@ function start_exareme(){ #Starts exareme daemon
-Djava.rmi.server.codebase=file:$EXAREME_HOME/lib/exareme/ \
-Djava.security.policy=$EXAREME_HOME/etc/exareme/art.policy \
-Djava.rmi.server.hostname=$EXAREME_MASTER \
-Dsun.rmi.activation.execTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dsun.rmi.activation.groupTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dsun.rmi.transport.connectionTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dsun.rmi.transport.proxy.connectTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dsun.rmi.transport.tcp.handshakeTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dsun.rmi.transport.tcp.responseTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dcom.sun.management.jmxremote.port=$EXAREME_ADMIN_JMX_PORT \
-Dcom.sun.management.jmxremote.authenticate=false \
-Dcom.sun.management.jmxremote.ssl=false \
Expand Down
4 changes: 4 additions & 0 deletions Exareme-Docker/files/root/exareme/start-worker.sh
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,10 @@ EXAREME_ADMIN_OPTS="${EXAREME_JAVA_OPTS} \
-Djava.rmi.server.codebase=file:$EXAREME_HOME/lib/exareme/ \
-Djava.security.policy=$EXAREME_HOME/etc/exareme/art.policy\
-Djava.rmi.server.hostname=$MY_IP \
-Dsun.rmi.activation.execTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dsun.rmi.activation.groupTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dsun.rmi.dgc.ackTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dsun.rmi.transport.tcp.readTimeout=$NODE_COMMUNICATION_TIMEOUT \
-Dcom.sun.management.jmxremote.port=$EXAREME_ADMIN_JMX_PORT\
-Dcom.sun.management.jmxremote.authenticate=false \
-Dcom.sun.management.jmxremote.ssl=false \
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -73,7 +73,7 @@ public void produceContent(ContentEncoder encoder, IOControl ioctrl)
log.trace("|" + queryStatus.getError() + "|");
if (queryStatus.getError().contains("ExaremeError:")) {
String data = queryStatus.getError().substring(queryStatus.getError().lastIndexOf("ExaremeError:") + "ExaremeError:".length()).replaceAll("\\s", " ");
//type could be error, user_error, warning regarding the error occured along the process
//type could be error, user_error, warning regarding the error occurred along the process
String type = user_error;
String result = defaultOutputFormat(data, type);
logErrorMessage(result);
Expand All @@ -82,7 +82,7 @@ public void produceContent(ContentEncoder encoder, IOControl ioctrl)
close();
} else if (queryStatus.getError().contains("PrivacyError")) {
String data = "The Experiment could not run with the input provided because there are insufficient data.";
//type could be error, user_error, warning regarding the error occured along the process
//type could be error, user_error, warning regarding the error occurred along the process
String type = warning;
String result = defaultOutputFormat(data, type);
logErrorMessage(result);
Expand All @@ -91,25 +91,25 @@ public void produceContent(ContentEncoder encoder, IOControl ioctrl)
close();
} else if (queryStatus.getError().contains("java.rmi.RemoteException")) {
String data = "One or more containers are not responding. Please inform the system administrator.";
//type could be error, user_error, warning regarding the error occured along the process
//type could be error, user_error, warning regarding the error occurred along the process
String type = error;
String result = defaultOutputFormat(data, type);
logErrorMessage(result);
encoder.write(ByteBuffer.wrap(result.getBytes()));
encoder.complete();
close();
} else if (queryStatus.getError().contains("java.lang.IndexOutOfBoundsException:")) {
String data = "Something went wrong. Clean-ups were made, you may re-run your experiment. Please inform the system administrator though for fixing any remaining issue.";
//type could be error, user_error, warning regarding the error occured along the process
String data = "Something went wrong. Please inform the system administrator.";
//type could be error, user_error, warning regarding the error occurred along the process
String type = error;
String result = defaultOutputFormat(data, type);
logErrorMessage(result);
encoder.write(ByteBuffer.wrap(result.getBytes()));
encoder.complete();
close();
} else {
String data = "Something went wrong. Please inform your system administrator to consult the logs.";
//type could be error, user_error, warning regarding the error occured along the process
String data = "Something went wrong. Please inform the system administrator.";
//type could be error, user_error, warning regarding the error occurred along the process
String type = error;
String result = defaultOutputFormat(data, type);
logErrorMessage(result);
Expand Down
3 changes: 2 additions & 1 deletion Federated-Deployment/Compose-Files/docker-compose-master.yml
Original file line number Diff line number Diff line change
Expand Up @@ -49,8 +49,9 @@ services:
- CONSULURL=${EXAREME_KEYSTORE}
- FEDERATION_ROLE=${FEDERATION_ROLE}
- NODE_NAME=${FEDERATION_NODE}
- ENVIRONMENT_TYPE=PROD #TEST / DEV / PROD
- TEMP_FILES_CLEANUP_TIME=30
- NODE_COMMUNICATION_TIMEOUT=30000 # (MILIS) NODE COMMUNICATION WILL DROP IF TIMEOUT IS PASSED
- ENVIRONMENT_TYPE=PROD # TEST / DEV / PROD
depends_on:
- exareme-keystore
deploy:
Expand Down
3 changes: 2 additions & 1 deletion Federated-Deployment/Compose-Files/docker-compose-worker.yml
Original file line number Diff line number Diff line change
Expand Up @@ -30,8 +30,9 @@ services:
- CONSULURL=${EXAREME_KEYSTORE}
- NODE_NAME=${FEDERATION_NODE}
- FEDERATION_ROLE=${FEDERATION_ROLE}
- ENVIRONMENT_TYPE=PROD #TEST / DEV / PROD
- TEMP_FILES_CLEANUP_TIME=30
- NODE_COMMUNICATION_TIMEOUT=30000 # (MILIS) NODE COMMUNICATION WILL DROP IF TIMEOUT IS PASSED
- ENVIRONMENT_TYPE=PROD # TEST / DEV / PROD
deploy:
restart_policy:
condition: on-failure
Expand Down
3 changes: 2 additions & 1 deletion Local-Deployment/docker-compose-master.yml
Original file line number Diff line number Diff line change
Expand Up @@ -48,8 +48,9 @@ services:
- CONSULURL=${EXAREME_KEYSTORE}
- FEDERATION_ROLE=${FEDERATION_ROLE}
- NODE_NAME=${FEDERATION_NODE}
- ENVIRONMENT_TYPE=PROD # TEST / DEV / PROD
- TEMP_FILES_CLEANUP_TIME=30
- NODE_COMMUNICATION_TIMEOUT=30000 # (MILIS) NODE COMMUNICATION WILL DROP IF TIMEOUT IS PASSED
- ENVIRONMENT_TYPE=PROD # TEST / DEV / PROD
depends_on:
- exareme-keystore
deploy:
Expand Down