fun emptyCache(parameters: Pair<FileSystem, Boolean>) {
    setUp(parameters.first, parameters.second)
    cache.close()
    assertJournalEquals()
  }

  @ParameterizedTest
  @ArgumentsSource(FileSystemParamProvider::class)
  fun recoverFromInitializationFailure(parameters: Pair<FileSystem, Boolean>) {
    setUp(parameters.first, parameters.second)

    mockWebServer: MockWebServer,
  ) {
    setUp(protocol, mockWebServer)
    recoverFromOneHttp2ErrorRequiresNewConnection(ErrorCode.INTERNAL_ERROR)
  }

  @ParameterizedTest
  @ArgumentsSource(ProtocolParamProvider::class)
  fun recoverFromOneCancelRequiresNewConnection(
    protocol: Protocol,
    mockWebServer: MockWebServer,
  ) {
    setUp(protocol, mockWebServer)

RUN /setup.python.sh python3.11 devel.requirements.txt
RUN /setup.python.sh python3.12 devel.requirements.txt

FROM devel as tf
# Setup TF Python environment.
COPY devel.requirements.txt /devel.requirements.txt
COPY setup.python.sh /setup.python.sh
RUN /setup.python.sh python3.9 devel.requirements.txt
RUN /setup.python.sh python3.10 devel.requirements.txt
RUN /setup.python.sh python3.11 devel.requirements.txt

**In our tests we also found ext4 does not honor POSIX O_DIRECT/Fdatasync semantics, ext4 trades performance for consistency guarantees. Please avoid ext4 in your setup.**

    enum class Type {
      Handshake,
      Plaintext,
      Encrypted,
      Setup,
      Unknown,
    }

    val type: Type
      get() =
        when {
          message == "adding as trusted certificates" -> Type.Setup
          message == "Raw read" || message == "Raw write" -> Type.Encrypted

is decided by how `domain.com` gets resolved, if there is a round-robin DNS on `domain.com` then
it is randomized which cluster might provision the bucket.

### 3. Test your setup

To test this setup, access the MinIO server via browser or [`mc`](https://min.io/docs/minio/linux/reference/minio-mc.html#quickstart). You’ll see the uploaded files are accessible from the all the MinIO endpoints.

## Explore Further

    - If excluded, ignore the pod and return prevResult
1. Setup redirect rules for the pods:
    - Get the port list from pods definition, as well as annotations.
    - Setup iptables with required port list: `nsenter --net=<k8s pod netns> /opt/cni/bin/istio-iptables ...`. Following conditions will prevent the redirect rules to be setup in the pods:

This allows the same configurations and lifecycle to apply to gateways as sidecars.

Note: this does mean that the namespace the gateway is deployed in must not have the `istio-injection=disabled` label.
See [Controlling the injection policy](https://istio.io/latest/docs/setup/additional-setup/sidecar-injection/#controlling-the-injection-policy) for more info.

### 2. Run MinIO Server with Erasure Code

Example: Start MinIO server in a 12 drives setup, using MinIO binary.

```sh
minio server /data{1...12}
```

Example: Start MinIO server in a 8 drives setup, using MinIO Docker image.

```sh
podman run \
  -p 9000:9000 \
  -p 9001:9001 \
  --name minio \
  -v /mnt/data1:/data1 \
  -v /mnt/data2:/data2 \

 | `application/x-compress` |
 | `application/x-xz`       |

All files with these extensions and mime types are excluded from compression,
even if compression is enabled for all types.

## To test the setup

To test this setup, practice put calls to the server using `mc` and use `mc ls` on
the data directory to view the size of the object.

## Explore Further