Bug 1626281

Summary: [3.9] fluentd pods are running with error logs which makes fill up disk very quickly.
Product: OpenShift Container Platform Reporter: Wang Haoran <haowang>
Component: LoggingAssignee: Rich Megginson <rmeggins>
Status: CLOSED ERRATA QA Contact: Anping Li <anli>
Severity: high Docs Contact:
Priority: high    
Version: 3.9.0CC: aos-bugs, jcantril, jmalde, rmeggins
Target Milestone: ---Keywords: OpsBlocker
Target Release: 3.9.z   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: openshift3/logging-fluentd:v3.9.44-1 Doc Type: Bug Fix
Doc Text:
Cause: Fluentd generates events internally for some unknown reason with the OneEventStream class. This class does not have the `empty?` method. The kubernetes metadata filter was using the `empty?` method on the EventStream object to avoid processing an empty stream. Consequence: Fluentd issues many error messages complaining about a missing `empty?` method. This overwhelms the container logging and causes disk issues. Fix: The kubernetes metadata filter was changed to only call the `empty?` method on objects which have that method. Result: fluentd logs should not be filled with this message.
Story Points: ---
Clone Of:
: 1626552 (view as bug list) Environment:
Last Closed: 2018-11-20 03:12:03 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1626552, 1626553    

Description Wang Haoran 2018-09-07 00:12:22 UTC
Description of problem:

2018-09-06 09:07:39 +0000 [warn]: emit transaction failed: error_class=NoMethodError error="undefined method `empty?' for #<Fluent::OneEventStream:0x007fbe0179ea70>" location="/usr/share/gems/gems/fluent-plugin-kubernetes_metadata_filter-1.0.3/lib/fluent/plugin/filter_kubernetes_metadata.rb:287:in `filter_stream_from_files'" tag="kubernetes.journal.container.fluentd"
  2018-09-06 09:07:39 +0000 [warn]: suppressed same stacktrace
2018-09-06 09:07:39 +0000 [warn]: emit transaction failed: error_class=NoMethodError error="undefined method `empty?' for #<Fluent::OneEventStream:0x007fbe0179ea70>" location="/usr/share/gems/gems/fluent-plugin-kubernetes_metadata_filter-1.0.3/lib/fluent/plugin/filter_kubernetes_metadata.rb:287:in `filter_stream_from_files'" tag="journal"
2018-09-06 09:07:39 +0000 [warn]: temporarily failed to flush the buffer. next_retry=2018-09-06 09:07:42 +0000 error_class="Fluent::ElasticsearchOutput::ConnectionFailure" error="Can not reach Elasticsearch cluster ({:host=>\"logging-es\", :port=>9200, :scheme=>\"https\"})!" plugin_id="object:3fdf005434e4"
  2018-09-06 09:07:39 +0000 [warn]: suppressed same stacktrace
  2018-09-06 09:07:39 +0000 [warn]: suppressed same stacktrace



Version-Release number of selected component (if applicable):

v3.9.40

How reproducible:


Steps to Reproduce:
1.
2.
3.

Actual results:


Expected results:


Additional info:

Comment 6 Anping Li 2018-10-15 05:54:44 UTC
Verified and pass with openshift3/logging-fluentd/images/v3.9.47-1

Comment 8 errata-xmlrpc 2018-11-20 03:12:03 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2018:2908