Blob CSI works outside of resource memory limits boundaries

[Vegoo89]

What happened:
Blob CSI doesn't respect memory limits when invoking blobfuse2 process on the host

What you expected to happen:
Tbh I don't know if its expected behaviour or not but I was sure that blobfuse2 is somehow respecting limits of blob pod in csi-blob-node daemonset. If this is expected there should be a big warning sign for using block cache, since by default it can use up to 80% of host free memory calculated on process start (so in most cases when nothing else is already allocated on the machine). Or maybe there is another way to work around it?

How to reproduce it:
Create any mount definition that uses Blob CSI driver with blobfuse2 version > 2.3.0 using block cache. Detailed parameters are described in azure-storage-fuse issue: Azure/azure-storage-fuse#1639

Anything else we need to know?:
None

Environment:

• CSI Driver version: 1.25.1
• Kubernetes version (use kubectl version): 1.30.7
• OS (e.g. from /etc/os-release): Ubuntu Linux (AKS)
• Kernel (e.g. uname -a): Ubuntu Linux (AKS)
• Install tools: helm chart
• Others: N/A