In the Talend Management Console Cloud production environment (Remote Engine v2.13.x), task executions are repeatedly failing with the error:
"Connection reset by peer`at `org.talend.remote.jobserver.client.JobSenderClient.sendJob".
Resolution
Increase the value of "MAX_ARCHIVES_DIR_SIZE" in RE/etc/org.talend.remote.jobserver.server.cfg to accommodate larger job archives and prevent further upload failures.
For example
Changing the following value from 100G to 300G, up to your requirement
org.talend.remote.jobserver.commons.config.JobServerConfiguration.MAX_ARCHIVES_DIR_SIZE=100G The default is 100 G
Based on interpretation of RE logs and data, the "Connection reset by peer "error originates from JobLimitException. Specifically, the "MAX_ARCHIVES_DIR_SIZE" limit defined in RE/etc/org.talend.remote.jobserver.server.cfg file has been reached, which causes the job upload request to be rejected on FILE_SERVER_PORT 8004.
The obvious JobLimitException error message looks like:
Caused by: org.talend.remote.jobserver.commons.utils.io.JobLimitException: Current upload of job archive will exceed the allowed size of archive directory of: '107374182400' (configured in org.talend.remote.jobserver.commons.config.JobServerConfiguration.MAX_ARCHIVES_DIR_SIZE)
Since the error "Connection reset by peer "message is not directly indicative of this root cause, it is necessary to search for relevant keywords in RE/data/log to determine whether any of the following job size limits have been exceeded: