Description of problem:

In case that not user defined values of limits.cpu and limits.memory,  Vector has not default values, then, as it's designed to work in memory, it can lead to get exhausted an OCP node (including master nodes).

The high memory usage is reported in https://issues.redhat.com/browse/LOG-4536 where Vector pods where able to use until 160GB of memory.

Then, it's required that in case that not user defined values for limits.cpu and limits.memory for the collector, the default installation should set some values for avoiding Vector consuming all the memory or cpu available on the nodes and impacting to the cluster

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

Logging 5.6, 5.7 and 5.8

How reproducible:

Always

Steps to Reproduce:

Deploy Vector without having user defined values and verify that the collector pods have not  _ limits.cpu_ or limits.memory values

Actual results:

If not user defined values of limits.cpu and limits.memory then, not default values assigned being not a secure and good configuration out of the box for the collector.

Expected results:

If not user defined values of limits.cpu and limits.memory, then, Vector collector can use all the memory and cpu from the nodes and it does it causing problems on the OCP nodes and the loads running on top of it.