From c9e92b47333d22a3cf47e76ac7248deb23263500 Mon Sep 17 00:00:00 2001 From: Rich Megginson Date: Thu, 6 Jul 2017 09:27:19 -0600 Subject: Bug 1466152 - Json-file log driver: Neither "openshift_logging_fluentd_use_journal=false" nor omitted collects the log entries https://bugzilla.redhat.com/show_bug.cgi?id=1466152 Do not set openshift_logging_fluentd_use_journal or USE_JOURNAL at all unless it is explicitly set as an ansible param. It is almost always better to let fluentd figure out which log driver docker is using. --- roles/openshift_hosted_logging/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'roles/openshift_hosted_logging/README.md') diff --git a/roles/openshift_hosted_logging/README.md b/roles/openshift_hosted_logging/README.md index 12ffe777d..680303853 100644 --- a/roles/openshift_hosted_logging/README.md +++ b/roles/openshift_hosted_logging/README.md @@ -22,7 +22,7 @@ - openshift_hosted_logging_kibana_nodeselector: Specify the nodeSelector that Kibana should be use (label=value) - openshift_hosted_logging_curator_nodeselector: Specify the nodeSelector that Curator should be use (label=value) - openshift_hosted_logging_enable_ops_cluster: If "true", configure a second ES cluster and Kibana for ops logs. -- openshift_hosted_logging_use_journal: If this is unset or empty, logging will try to figure out from docker which log driver it is using (json-file or journald). You can set this param to "true" or "false" to force logging to use journal or not (but make sure you are sure which one docker is using). +- openshift_hosted_logging_use_journal: *DEPRECATED - DO NOT USE* - openshift_hosted_logging_journal_source: By default, if this param is unset or empty, logging will use `/var/log/journal` if it exists, or `/run/log/journal` if not. You can use this param to force logging to use a different location. - openshift_hosted_logging_journal_read_from_head: Set to `true` to have fluentd read from the beginning of the journal, to get historical log data. Default is `false`. *WARNING* Using `true` may take several minutes or even hours, depending on the size of the journal, until any new records show up in Elasticsearch, and will cause fluentd to consume a lot of CPU and RAM resources. -- cgit v1.2.3