Description of problem: This bug seems to be solved on 3.9: https://bugzilla.redhat.com/show_bug.cgi?id=1562004 ( errata: https://access.redhat.com/errata/RHBA-2018:2335 ) Customer is using ocp v3.6 and has tha same issue: 2018-09-03 15:10:12 -0300 [warn]: temporarily failed to flush the buffer. next_retry=2018-09-03 15:15:10 -0300 error_class="Elasticsearch::Transport::Transport::Errors::InternalServerError" error="[500] {\"error\":{\"root_cause\":[{\"type\":\"json_parse_exception\",\"reason\":\"Invalid UTF-8 start byte 0x92\\n at [Source: [B@12278e45; line: 1, column: 106]\"}],\"type\":\"json_parse_exception\",\"reason\":\"Invalid UTF-8 start byte 0x92\\n at [Source: [B@12278e45; line: 1, column: 106]\"},\"status\":500}" plugin_id="object:b0df74" I would like to request the analysis of a backport for 3.6. Regards
Actually, I believe we backported this change into 3.7 along with numerous other fixes to fluent to make it significantly more stable.
Correct
Closing as no additional information for some time.
(In reply to Jeff Cantrill from comment #25) > Closing as no additional information for some time. Ok, will let you know in case the issue persists. Thanks for the help.
Closing WONTFIX. Current open issue: https://bugzilla.redhat.com/show_bug.cgi?id=1654704