should be able to set priorityClassName for fluent

[brianmay]

My observation is unless I can set priorityClassName: system-node-critical for the fluent-bit daemon set, the cluster can end up in a broken state where it can't start fluent-bit on a node because there are not enough resources on a given node, but it will not kick off any of the pods that are consuming the resources either.

By setting this priority, it will kick off some of the less important pods, which will be rescheduled on other nodes, and fluent-bit is allowed to run.

But this chart doesn't appear to allow setting that value.

[arenstar]

Second that - this chart requires priorityClassName

[prem-pdmi]

When this will be released?