[JIRA] (JENKINS-60737) EC2 ASG Slave taken offline in middle of a job for no diskspace left when multiple executors given to a slave

[arindom.sarkar@gmail.com (JIRA)]

Arindom Sarkar created an issue

Jenkins / JENKINS-60737 EC2 ASG Slave taken offline in middle of a job for no diskspace left when multiple executors given to a slave Issue Type: Bug Assignee: Chad Schmutzer Components: ec2-fleet-plugin Created: 2020-01-10 22:38 Priority: Minor Reporter: Arindom Sarkar slaves spawned off by EC2 fleet plugin were taken offline with this message. This happens when multiple executors assigned to the slave, and jobs are allotted to them concurrently and every jobs checkout a git repository. Is there a way to avoid this? 13:31:12 nonprodslave2 i-0dd32************ was marked offline: Node is being removed Receiving objects: 79% (88346/111053), 179.52 MiB | 2.03 MiB/s fatal: write error: No space left on device*13:30:29* fatal: index-pack failed*13:30:29* Add Comment

This message was sent by Atlassian Jira (v7.13.6#713006-sha1:cc4451f)

[artem.stasuk@gmail.com (JIRA)]

Artem Stasiuk assigned an issue to Artem Stasiuk

Jenkins / JENKINS-60737 EC2 ASG Slave taken offline in middle of a job for no diskspace left when multiple executors given to a slave

Change By: Artem Stasiuk Assignee: Chad Schmutzer Artem Stasiuk

Add Comment

[artem.stasuk@gmail.com (JIRA)]

Artem Stasiuk updated JENKINS-60737

Jenkins / JENKINS-60737 EC2 ASG Slave taken offline in middle of a job for no diskspace left when multiple executors given to a slave

Change By: Artem Stasiuk Status: Open Fixed but Unreleased Resolution: Not A Defect

Add Comment

[artem.stasuk@gmail.com (JIRA)]

Artem Stasiuk updated JENKINS-60737

Jenkins / JENKINS-60737 EC2 ASG Slave taken offline in middle of a job for no diskspace left when multiple executors given to a slave

Change By: Artem Stasiuk Status: Fixed but Unreleased Resolved

Add Comment

[artem.stasuk@gmail.com (JIRA)]

Artem Stasiuk commented on JENKINS-60737

Re: EC2 ASG Slave taken offline in middle of a job for no diskspace left when multiple executors given to a slave Doesn't seem like a bug, looks like builds consume a lot of space, you can try to reduce amount of parallel builds and make sure you checkout only files without repo history https://stackoverflow.com/questions/29368837/copy-a-git-repo-without-history

Add Comment