bug 1590920. Bump fluent default memory to 756M

[jcantrill]

This resolves https://bugzilla.redhat.com/show_bug.cgi?id=1590920 by bumping the default OOTB memory for fluent to support prometheus metrics and buffer retry q's

[jcantrill]

@mffiedler Is your opinion we should backport this to 3.10?

[mffiedler]

@jcantrill IMO yes a backport to 3.10 would be good. 3.10 steady state fluentd rss is up to 50% higher than 3.9, in part depending on how much buffering and retries are taking place. Functionally, this seems to be the tradeoff for the greatly improved reduction in duplicates and message loss. The higher memory limit should help reduce oom kills and restarts for the workload I was running (500 1K messages/second/fluentd).

Other workloads might still require tuning the value further.

[jcantrill]

/test logging

[ewolinetz]

bot, retest this please

[ewolinetz]

/test logging

[ewolinetz]

/lgtm

[openshift-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: ewolinetz, jcantrill

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• roles/openshift_logging_fluentd/OWNERS [ewolinetz,jcantrill]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[richm]

/test logging

[jcantrill]

/cherrypick release-3.10

[openshift-cherrypick-robot]

@jcantrill: new pull request created: #9191

In response to this:

/cherrypick release-3.10

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.