LimitRange are automatically applied to all the pods that don't define resources (requests/limits). Currently, you can only set memory limits for some Syndesis components. The most heavy component is syndesis-server which, by default, requires 450m of CPU with a limit of 750m.
Now, if you are on a shared cluster with LimitRange, you are forced to set a CPU request which is equal to the most heavy pod (component or integration), but this is then requested by all pods. We need a way to set specific CPU/mem values for all pods created by Syndesis at the CR level and also a way to set specific resources for some heavy integrations.