[STORM-3319] Slot can fail assertions in some cases

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.0.0
• Fix Version/s: 2.0.0
• Labels: pull-request-available
• Assignee: Stig Rohde Døssing
• Reporter: Stig Rohde Døssing
• Created: 21/Jan/19 12:43
• Updated: 16/Mar/19 03:26
• Resolved: 15/Mar/19 18:44

Description

2019-01-19 22:47:03.045 [SLOT_1024] ERROR org.apache.storm.daemon.supervisor.Slot - Error when processing event
java.lang.AssertionError: null
at org.apache.storm.daemon.supervisor.Slot.handleEmpty(Slot.java:781) ~[classes/:?]
at org.apache.storm.daemon.supervisor.Slot.stateMachineStep(Slot.java:217) ~[classes/:?]
at org.apache.storm.daemon.supervisor.Slot.run(Slot.java:900) [classes/:?]
2019-01-19 22:47:03.045 [SLOT_1025] ERROR org.apache.storm.daemon.supervisor.Slot - Error when processing event
java.lang.AssertionError: null
at org.apache.storm.daemon.supervisor.Slot.handleEmpty(Slot.java:781) ~[classes/:?]
at org.apache.storm.daemon.supervisor.Slot.stateMachineStep(Slot.java:217) ~[classes/:?]
at org.apache.storm.daemon.supervisor.Slot.run(Slot.java:900) [classes/:?]

The issue is that Slot tries to go from WAITING_FOR_LOCALIZATION to EMPTY when there's an exception downloading a blob. It then fails one of the assertions in EMPTY because it doesn't clear its pendingChangingBlobsAssignment field.

There's no reason to go back to EMPTY. The Slot still wants to download some blobs, so it should just restart the downloads and go back to WAITING_FOR_LOCALIZATION.

Issue Links

• links to: GitHub Pull Request #2943