# HugePage

## HugePage

HugePage is a new feature introduced in v1.9 (v1.9 Alpha, v1.10 Beta) that enables direct use of HugePages on a Node within containers.

### Configuration

* Enable the feature by setting `--feature-gates=HugePages=true`
* Pre-allocate HugePages on the Node with commands like:

```bash
mount -t hugetlbfs \
    -o uid=<value>,gid=<value>,mode=<value>,pagesize=<value>,size=<value>,\
    min_size=<value>,nr_inodes=<value> none /mnt/huge
```

### Usage

Here's a sample configuration of a Pod that uses HugePages:

```yaml
apiVersion: v1
kind: Pod
metadata:
  generateName: hugepages-volume-
spec:
  containers:
  - image: fedora:latest
    command:
    - sleep
    - inf
    name: example
    volumeMounts:
    - mountPath: /hugepages
      name: hugepage
    resources:
      limits:
        hugepages-2Mi: 100Mi
  volumes:
  - name: hugepage
    emptyDir:
      medium: HugePages
```

Please Note:

* The request for HugePages and the limit must be the same.
* HugePage provides Pod-level isolation, but not yet at the container level.
* EmptyDir volumes based on HugePage can only use the HugePage memory requested.
* Usage of HugePages can be restricted through ResourceQuota.
* When obtaining HugePages within a container application using `shmget(SHM_HUGETLB)`, it is essential to configure the user group to match the one in `proc/sys/vm/hugetlb_shm_group` (`securityContext.SupplementalGroups`).

***

## HugePages Unleashed

Unlocking the Power of Massive Memory Pages in Containers

### Setting the Stage for HugePages

HugePage is a savvy innovation ushered in with v1.9 (Alpha in v1.9, Beta in v1.10), designed to let containers harness the might of Node-level HugePages without a hitch.

### Config Sheet for the Tech-Savvy

To get started with HugePages:

* Flick on the HugePages feature with a simple flag: `--feature-gates=HugePages=true`
* Next, line up your HugePages on the Node like ducks in a row with a command akin to:

```bash
mount -t hugetlbfs \
    -o uid=<value>,gid=<value>,mode=<value>,pagesize=<value>,size=<value>,\
    min_size=<value>,nr_inodes=<value> none /mnt/huge
```

### Entering the HugePages Era

Craft your Pod with a flair for the huge—here's a blueprint to get you rolling:

```yaml
apiVersion: v1
kind: Pod
metadata:
  generateName: hugepages-volume-
spec:
  containers:
  - image: fedora:latest
    command:
    - sleep
    - inf
    name: example
    volumeMounts:
    - mountPath: /hugepages
      name: hugepage
    resources:
      limits:
        hugepages-2Mi: 100Mi
  volumes:
  - name: hugepage
    emptyDir:
      medium: HugePages
```

A Few Pro Tips:

* When dealing with HugePages, make sure your requests match your limits.
* Think of HugePages as a Pod-exclusive club—no container can crash this party alone.
* Your EmptyDir drawers will be custom-fitted just for those HugePages you've asked for.
* Keep a leash on those HugePages with wise ResourceQuota policies.
* Want to snag HugePages in your container's app? Sync up with the user group vibe set by `proc/sys/vm/hugetlb_shm_group` by tweaking `securityContext.SupplementalGroups`.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://kubernetes.feisky.xyz/en/practice/hugepage.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.