[MESOS-8871] Agent may fail to recover if the agent dies before image store cache checkpointed.

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.4.2, 1.5.2, 1.6.1, 1.7.0
• Component/s: agent
• Labels: mesosphere slave
• Epic Link: Unified Container
• Sprint: Mesosphere Sprint 2018-23
• Story Points: 3
• Assignee: bin zheng
• Reporter: Gilbert Song
• Shepherd: Gilbert Song
• Created: 02/May/18 16:58
• Updated: 04/Jul/18 01:23
• Resolved: 21/Jun/18 20:18

Description

E0502 13:51:45.398555 10100 slave.cpp:7305] EXIT with status 1: Failed to perform recovery: Collect failed: Collect failed: Collect failed: Unexpected empty images file '/var/lib/mesos/slave/store/docker/storedImages'

This may happen if the agent dies after the file is created but before the contents are persisted on disk.