

기계 번역으로 제공되는 번역입니다. 제공된 번역과 원본 영어의 내용이 상충하는 경우에는 영어 버전이 우선합니다.

# SDK for Rust를 사용한 Amazon EC2 예제
<a name="rust_ec2_code_examples"></a>

다음 코드 예제에서는 AWS SDK for Rust를 Amazon EC2와 함께 사용하여 작업을 수행하고 일반적인 시나리오를 구현하는 방법을 보여줍니다.

*기본 사항*은 서비스 내에서 필수 작업을 수행하는 방법을 보여주는 코드 예제입니다.

*작업*은 대규모 프로그램에서 발췌한 코드이며 컨텍스트에 맞춰 실행해야 합니다. 작업은 개별 서비스 함수를 직접적으로 호출하는 방법을 보여주며 관련 시나리오의 컨텍스트에 맞는 작업을 볼 수 있습니다.

각 예시에는 전체 소스 코드에 대한 링크가 포함되어 있으며, 여기에서 컨텍스트에 맞춰 코드를 설정하고 실행하는 방법에 대한 지침을 찾을 수 있습니다.

**Topics**
+ [시작하기](#get_started)
+ [기본 사항](#basics)
+ [작업](#actions)

## 시작하기
<a name="get_started"></a>

### Hello Amazon EC2
<a name="ec2_Hello_rust_topic"></a>

다음 코드 예제에서는 Amazon EC2 사용을 시작하는 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
async fn show_security_groups(client: &aws_sdk_ec2::Client, group_ids: Vec<String>) {
    let response = client
        .describe_security_groups()
        .set_group_ids(Some(group_ids))
        .send()
        .await;

    match response {
        Ok(output) => {
            for group in output.security_groups() {
                println!(
                    "Found Security Group {} ({}), vpc id {} and description {}",
                    group.group_name().unwrap_or("unknown"),
                    group.group_id().unwrap_or("id-unknown"),
                    group.vpc_id().unwrap_or("vpcid-unknown"),
                    group.description().unwrap_or("(none)")
                );
            }
        }
        Err(err) => {
            let err = err.into_service_error();
            let meta = err.meta();
            let message = meta.message().unwrap_or("unknown");
            let code = meta.code().unwrap_or("unknown");
            eprintln!("Error listing EC2 Security Groups: ({code}) {message}");
        }
    }
}
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DescribeSecurityGroups](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_security_groups)을 참조하세요.

## 기본 사항
<a name="basics"></a>

### 기본 사항 알아보기
<a name="ec2_Scenario_GetStartedInstances_rust_topic"></a>

다음 코드 예제에서는 다음과 같은 작업을 수행하는 방법을 보여줍니다.
+ 키 페어 및 보안 그룹을 생성합니다.
+ Amazon Machine Image(AMI) 및 호환되는 인스턴스 유형을 선택한 다음 인스턴스를 생성합니다.
+ 인스턴스를 중지한 후 다시 시작합니다.
+ 인스턴스와 탄력적 IP 주소 연결.
+ SSH로 인스턴스에 연결한 다음 리소스를 정리합니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.
EC2InstanceScenario 구현에는 예제를 전체적으로 실행하는 로직이 포함되어 있습니다.  

```
//! Scenario that uses the AWS SDK for Rust (the SDK) with Amazon Elastic Compute Cloud
//! (Amazon EC2) to do the following:
//!
//! * Create a key pair that is used to secure SSH communication between your computer and
//!   an EC2 instance.
//! * Create a security group that acts as a virtual firewall for your EC2 instances to
//!   control incoming and outgoing traffic.
//! * Find an Amazon Machine Image (AMI) and a compatible instance type.
//! * Create an instance that is created from the instance type and AMI you select, and
//!   is configured to use the security group and key pair created in this example.
//! * Stop and restart the instance.
//! * Create an Elastic IP address and associate it as a consistent IP address for your instance.
//! * Connect to your instance with SSH, using both its public IP address and your Elastic IP
//!   address.
//! * Clean up all of the resources created by this example.

use std::net::Ipv4Addr;

use crate::{
    ec2::{EC2Error, EC2},
    getting_started::{key_pair::KeyPairManager, util::Util},
    ssm::SSM,
};
use aws_sdk_ssm::types::Parameter;

use super::{
    elastic_ip::ElasticIpManager, instance::InstanceManager, security_group::SecurityGroupManager,
    util::ScenarioImage,
};

pub struct Ec2InstanceScenario {
    ec2: EC2,
    ssm: SSM,
    util: Util,
    key_pair_manager: KeyPairManager,
    security_group_manager: SecurityGroupManager,
    instance_manager: InstanceManager,
    elastic_ip_manager: ElasticIpManager,
}

impl Ec2InstanceScenario {
    pub fn new(ec2: EC2, ssm: SSM, util: Util) -> Self {
        Ec2InstanceScenario {
            ec2,
            ssm,
            util,
            key_pair_manager: Default::default(),
            security_group_manager: Default::default(),
            instance_manager: Default::default(),
            elastic_ip_manager: Default::default(),
        }
    }

    pub async fn run(&mut self) -> Result<(), EC2Error> {
        self.create_and_list_key_pairs().await?;
        self.create_security_group().await?;
        self.create_instance().await?;
        self.stop_and_start_instance().await?;
        self.associate_elastic_ip().await?;
        self.stop_and_start_instance().await?;
        Ok(())
    }

    /// 1. Creates an RSA key pair and saves its private key data as a .pem file in secure
    ///    temporary storage. The private key data is deleted after the example completes.
    /// 2. Optionally, lists the first five key pairs for the current account.
    pub async fn create_and_list_key_pairs(&mut self) -> Result<(), EC2Error> {
        println!( "Let's create an RSA key pair that you can be use to securely connect to your EC2 instance.");

        let key_name = self.util.prompt_key_name()?;

        self.key_pair_manager
            .create(&self.ec2, &self.util, key_name)
            .await?;

        println!(
            "Created a key pair {} and saved the private key to {:?}.",
            self.key_pair_manager
                .key_pair()
                .key_name()
                .ok_or_else(|| EC2Error::new("No key name after creating key"))?,
            self.key_pair_manager
                .key_file_path()
                .ok_or_else(|| EC2Error::new("No key file after creating key"))?
        );

        if self.util.should_list_key_pairs()? {
            for pair in self.key_pair_manager.list(&self.ec2).await? {
                println!(
                    "Found {:?} key {} with fingerprint:\t{:?}",
                    pair.key_type(),
                    pair.key_name().unwrap_or("Unknown"),
                    pair.key_fingerprint()
                );
            }
        }

        Ok(())
    }

    /// 1. Creates a security group for the default VPC.
    /// 2. Adds an inbound rule to allow SSH. The SSH rule allows only
    ///    inbound traffic from the current computer’s public IPv4 address.
    /// 3. Displays information about the security group.
    ///
    /// This function uses <http://checkip.amazonaws.com> to get the current public IP
    /// address of the computer that is running the example. This method works in most
    /// cases. However, depending on how your computer connects to the internet, you
    /// might have to manually add your public IP address to the security group by using
    /// the AWS Management Console.
    pub async fn create_security_group(&mut self) -> Result<(), EC2Error> {
        println!("Let's create a security group to manage access to your instance.");
        let group_name = self.util.prompt_security_group_name()?;

        self.security_group_manager
            .create(
                &self.ec2,
                &group_name,
                "Security group for example: get started with instances.",
            )
            .await?;

        println!(
            "Created security group {} in your default VPC {}.",
            self.security_group_manager.group_name(),
            self.security_group_manager
                .vpc_id()
                .unwrap_or("(unknown vpc)")
        );

        let check_ip = self.util.do_get("https://checkip.amazonaws.com").await?;
        let current_ip_address: Ipv4Addr = check_ip.trim().parse().map_err(|e| {
            EC2Error::new(format!(
                "Failed to convert response {} to IP Address: {e:?}",
                check_ip
            ))
        })?;

        println!("Your public IP address seems to be {current_ip_address}");
        if self.util.should_add_to_security_group() {
            match self
                .security_group_manager
                .authorize_ingress(&self.ec2, current_ip_address)
                .await
            {
                Ok(_) => println!("Security group rules updated"),
                Err(err) => eprintln!("Couldn't update security group rules: {err:?}"),
            }
        }
        println!("{}", self.security_group_manager);

        Ok(())
    }

    /// 1. Gets a list of Amazon Linux 2 AMIs from AWS Systems Manager. Specifying the
    ///    '/aws/service/ami-amazon-linux-latest' path returns only the latest AMIs.
    /// 2. Gets and displays information about the available AMIs and lets you select one.
    /// 3. Gets a list of instance types that are compatible with the selected AMI and
    ///    lets you select one.
    /// 4. Creates an instance with the previously created key pair and security group,
    ///    and the selected AMI and instance type.
    /// 5. Waits for the instance to be running and then displays its information.
    pub async fn create_instance(&mut self) -> Result<(), EC2Error> {
        let ami = self.find_image().await?;

        let instance_types = self
            .ec2
            .list_instance_types(&ami.0)
            .await
            .map_err(|e| e.add_message("Could not find instance types"))?;
        println!(
            "There are several instance types that support the {} architecture of the image.",
            ami.0
                .architecture
                .as_ref()
                .ok_or_else(|| EC2Error::new(format!("Missing architecture in {:?}", ami.0)))?
        );
        let instance_type = self.util.select_instance_type(instance_types)?;

        println!("Creating your instance and waiting for it to start...");
        self.instance_manager
            .create(
                &self.ec2,
                ami.0
                    .image_id()
                    .ok_or_else(|| EC2Error::new("Could not find image ID"))?,
                instance_type,
                self.key_pair_manager.key_pair(),
                self.security_group_manager
                    .security_group()
                    .map(|sg| vec![sg])
                    .ok_or_else(|| EC2Error::new("Could not find security group"))?,
            )
            .await
            .map_err(|e| e.add_message("Scenario failed to create instance"))?;

        while let Err(err) = self
            .ec2
            .wait_for_instance_ready(self.instance_manager.instance_id(), None)
            .await
        {
            println!("{err}");
            if !self.util.should_continue_waiting() {
                return Err(err);
            }
        }

        println!("Your instance is ready:\n{}", self.instance_manager);

        self.display_ssh_info();

        Ok(())
    }

    async fn find_image(&mut self) -> Result<ScenarioImage, EC2Error> {
        let params: Vec<Parameter> = self
            .ssm
            .list_path("/aws/service/ami-amazon-linux-latest")
            .await
            .map_err(|e| e.add_message("Could not find parameters for available images"))?
            .into_iter()
            .filter(|param| param.name().is_some_and(|name| name.contains("amzn2")))
            .collect();
        let amzn2_images: Vec<ScenarioImage> = self
            .ec2
            .list_images(params)
            .await
            .map_err(|e| e.add_message("Could not find images"))?
            .into_iter()
            .map(ScenarioImage::from)
            .collect();
        println!("We will now create an instance from an Amazon Linux 2 AMI");
        let ami = self.util.select_scenario_image(amzn2_images)?;
        Ok(ami)
    }

    // 1. Stops the instance and waits for it to stop.
    // 2. Starts the instance and waits for it to start.
    // 3. Displays information about the instance.
    // 4. Displays an SSH connection string. When an Elastic IP address is associated
    //    with the instance, the IP address stays consistent when the instance stops
    //    and starts.
    pub async fn stop_and_start_instance(&self) -> Result<(), EC2Error> {
        println!("Let's stop and start your instance to see what changes.");
        println!("Stopping your instance and waiting until it's stopped...");
        self.instance_manager.stop(&self.ec2).await?;
        println!("Your instance is stopped. Restarting...");
        self.instance_manager.start(&self.ec2).await?;
        println!("Your instance is running.");
        println!("{}", self.instance_manager);
        if self.elastic_ip_manager.public_ip() == "0.0.0.0" {
            println!("Every time your instance is restarted, its public IP address changes.");
        } else {
            println!(
                "Because you have associated an Elastic IP with your instance, you can connect by using a consistent IP address after the instance restarts."
            );
        }
        self.display_ssh_info();
        Ok(())
    }

    /// 1. Allocates an Elastic IP address and associates it with the instance.
    /// 2. Displays an SSH connection string that uses the Elastic IP address.
    async fn associate_elastic_ip(&mut self) -> Result<(), EC2Error> {
        self.elastic_ip_manager.allocate(&self.ec2).await?;
        println!(
            "Allocated static Elastic IP address: {}",
            self.elastic_ip_manager.public_ip()
        );

        self.elastic_ip_manager
            .associate(&self.ec2, self.instance_manager.instance_id())
            .await?;
        println!("Associated your Elastic IP with your instance.");
        println!("You can now use SSH to connect to your instance by using the Elastic IP.");
        self.display_ssh_info();
        Ok(())
    }

    /// Displays an SSH connection string that can be used to connect to a running
    /// instance.
    fn display_ssh_info(&self) {
        let ip_addr = if self.elastic_ip_manager.has_allocation() {
            self.elastic_ip_manager.public_ip()
        } else {
            self.instance_manager.instance_ip()
        };
        let key_file_path = self.key_pair_manager.key_file_path().unwrap();
        println!("To connect, open another command prompt and run the following command:");
        println!("\nssh -i {} ec2-user@{ip_addr}\n", key_file_path.display());
        let _ = self.util.enter_to_continue();
    }

    /// 1. Disassociate and delete the previously created Elastic IP.
    /// 2. Terminate the previously created instance.
    /// 3. Delete the previously created security group.
    /// 4. Delete the previously created key pair.
    pub async fn clean_up(self) {
        println!("Let's clean everything up. This example created these resources:");
        println!(
            "\tKey pair: {}",
            self.key_pair_manager
                .key_pair()
                .key_name()
                .unwrap_or("(unknown key pair)")
        );
        println!(
            "\tSecurity group: {}",
            self.security_group_manager.group_name()
        );
        println!(
            "\tInstance: {}",
            self.instance_manager.instance_display_name()
        );
        if self.util.should_clean_resources() {
            if let Err(err) = self.elastic_ip_manager.remove(&self.ec2).await {
                eprintln!("{err}")
            }
            if let Err(err) = self.instance_manager.delete(&self.ec2).await {
                eprintln!("{err}")
            }
            if let Err(err) = self.security_group_manager.delete(&self.ec2).await {
                eprintln!("{err}");
            }
            if let Err(err) = self.key_pair_manager.delete(&self.ec2, &self.util).await {
                eprintln!("{err}");
            }
        } else {
            println!("Ok, not cleaning up any resources!");
        }
    }
}

pub async fn run(mut scenario: Ec2InstanceScenario) {
    println!("--------------------------------------------------------------------------------");
    println!(
        "Welcome to the Amazon Elastic Compute Cloud (Amazon EC2) get started with instances demo."
    );
    println!("--------------------------------------------------------------------------------");

    if let Err(err) = scenario.run().await {
        eprintln!("There was an error running the scenario: {err}")
    }

    println!("--------------------------------------------------------------------------------");

    scenario.clean_up().await;

    println!("Thanks for running!");
    println!("--------------------------------------------------------------------------------");
}
```
EC2Impl 구조체는 테스트를 위한 자동 모의점 역할을 하며 해당 함수는 EC2 SDK 직접 호출을 래핑합니다.  

```
use std::{net::Ipv4Addr, time::Duration};

use aws_sdk_ec2::{
    client::Waiters,
    error::ProvideErrorMetadata,
    operation::{
        allocate_address::AllocateAddressOutput, associate_address::AssociateAddressOutput,
    },
    types::{
        DomainType, Filter, Image, Instance, InstanceType, IpPermission, IpRange, KeyPairInfo,
        SecurityGroup, Tag,
    },
    Client as EC2Client,
};
use aws_sdk_ssm::types::Parameter;
use aws_smithy_runtime_api::client::waiters::error::WaiterError;

#[cfg(test)]
use mockall::automock;

#[cfg(not(test))]
pub use EC2Impl as EC2;

#[cfg(test)]
pub use MockEC2Impl as EC2;

#[derive(Clone)]
pub struct EC2Impl {
    pub client: EC2Client,
}

#[cfg_attr(test, automock)]
impl EC2Impl {
    pub fn new(client: EC2Client) -> Self {
        EC2Impl { client }
    }

    pub async fn create_key_pair(&self, name: String) -> Result<(KeyPairInfo, String), EC2Error> {
        tracing::info!("Creating key pair {name}");
        let output = self.client.create_key_pair().key_name(name).send().await?;
        let info = KeyPairInfo::builder()
            .set_key_name(output.key_name)
            .set_key_fingerprint(output.key_fingerprint)
            .set_key_pair_id(output.key_pair_id)
            .build();
        let material = output
            .key_material
            .ok_or_else(|| EC2Error::new("Create Key Pair has no key material"))?;
        Ok((info, material))
    }

    pub async fn list_key_pair(&self) -> Result<Vec<KeyPairInfo>, EC2Error> {
        let output = self.client.describe_key_pairs().send().await?;
        Ok(output.key_pairs.unwrap_or_default())
    }

    pub async fn delete_key_pair(&self, key_name: &str) -> Result<(), EC2Error> {
        let key_name: String = key_name.into();
        tracing::info!("Deleting key pair {key_name}");
        self.client
            .delete_key_pair()
            .key_name(key_name)
            .send()
            .await?;
        Ok(())
    }

    pub async fn create_security_group(
        &self,
        name: &str,
        description: &str,
    ) -> Result<SecurityGroup, EC2Error> {
        tracing::info!("Creating security group {name}");
        let create_output = self
            .client
            .create_security_group()
            .group_name(name)
            .description(description)
            .send()
            .await
            .map_err(EC2Error::from)?;

        let group_id = create_output
            .group_id
            .ok_or_else(|| EC2Error::new("Missing security group id after creation"))?;

        let group = self
            .describe_security_group(&group_id)
            .await?
            .ok_or_else(|| {
                EC2Error::new(format!("Could not find security group with id {group_id}"))
            })?;

        tracing::info!("Created security group {name} as {group_id}");

        Ok(group)
    }

    /// Find a single security group, by ID. Returns Err if multiple groups are found.
    pub async fn describe_security_group(
        &self,
        group_id: &str,
    ) -> Result<Option<SecurityGroup>, EC2Error> {
        let group_id: String = group_id.into();
        let describe_output = self
            .client
            .describe_security_groups()
            .group_ids(&group_id)
            .send()
            .await?;

        let mut groups = describe_output.security_groups.unwrap_or_default();

        match groups.len() {
            0 => Ok(None),
            1 => Ok(Some(groups.remove(0))),
            _ => Err(EC2Error::new(format!(
                "Expected single group for {group_id}"
            ))),
        }
    }

    /// Add an ingress rule to a security group explicitly allowing IPv4 address
    /// as {ip}/32 over TCP port 22.
    pub async fn authorize_security_group_ssh_ingress(
        &self,
        group_id: &str,
        ingress_ips: Vec<Ipv4Addr>,
    ) -> Result<(), EC2Error> {
        tracing::info!("Authorizing ingress for security group {group_id}");
        self.client
            .authorize_security_group_ingress()
            .group_id(group_id)
            .set_ip_permissions(Some(
                ingress_ips
                    .into_iter()
                    .map(|ip| {
                        IpPermission::builder()
                            .ip_protocol("tcp")
                            .from_port(22)
                            .to_port(22)
                            .ip_ranges(IpRange::builder().cidr_ip(format!("{ip}/32")).build())
                            .build()
                    })
                    .collect(),
            ))
            .send()
            .await?;
        Ok(())
    }

    pub async fn delete_security_group(&self, group_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Deleting security group {group_id}");
        self.client
            .delete_security_group()
            .group_id(group_id)
            .send()
            .await?;
        Ok(())
    }

    pub async fn list_images(&self, ids: Vec<Parameter>) -> Result<Vec<Image>, EC2Error> {
        let image_ids = ids.into_iter().filter_map(|p| p.value).collect();
        let output = self
            .client
            .describe_images()
            .set_image_ids(Some(image_ids))
            .send()
            .await?;

        let images = output.images.unwrap_or_default();
        if images.is_empty() {
            Err(EC2Error::new("No images for selected AMIs"))
        } else {
            Ok(images)
        }
    }

    /// List instance types that match an image's architecture and are free tier eligible.
    pub async fn list_instance_types(&self, image: &Image) -> Result<Vec<InstanceType>, EC2Error> {
        let architecture = format!(
            "{}",
            image.architecture().ok_or_else(|| EC2Error::new(format!(
                "Image {:?} does not have a listed architecture",
                image.image_id()
            )))?
        );
        let free_tier_eligible_filter = Filter::builder()
            .name("free-tier-eligible")
            .values("false")
            .build();
        let supported_architecture_filter = Filter::builder()
            .name("processor-info.supported-architecture")
            .values(architecture)
            .build();
        let response = self
            .client
            .describe_instance_types()
            .filters(free_tier_eligible_filter)
            .filters(supported_architecture_filter)
            .send()
            .await?;

        Ok(response
            .instance_types
            .unwrap_or_default()
            .into_iter()
            .filter_map(|iti| iti.instance_type)
            .collect())
    }

    pub async fn create_instance<'a>(
        &self,
        image_id: &'a str,
        instance_type: InstanceType,
        key_pair: &'a KeyPairInfo,
        security_groups: Vec<&'a SecurityGroup>,
    ) -> Result<String, EC2Error> {
        let run_instances = self
            .client
            .run_instances()
            .image_id(image_id)
            .instance_type(instance_type)
            .key_name(
                key_pair
                    .key_name()
                    .ok_or_else(|| EC2Error::new("Missing key name when launching instance"))?,
            )
            .set_security_group_ids(Some(
                security_groups
                    .iter()
                    .filter_map(|sg| sg.group_id.clone())
                    .collect(),
            ))
            .min_count(1)
            .max_count(1)
            .send()
            .await?;

        if run_instances.instances().is_empty() {
            return Err(EC2Error::new("Failed to create instance"));
        }

        let instance_id = run_instances.instances()[0].instance_id().unwrap();
        let response = self
            .client
            .create_tags()
            .resources(instance_id)
            .tags(
                Tag::builder()
                    .key("Name")
                    .value("From SDK Examples")
                    .build(),
            )
            .send()
            .await;

        match response {
            Ok(_) => tracing::info!("Created {instance_id} and applied tags."),
            Err(err) => {
                tracing::info!("Error applying tags to {instance_id}: {err:?}");
                return Err(err.into());
            }
        }

        tracing::info!("Instance is created.");

        Ok(instance_id.to_string())
    }

    /// Wait for an instance to be ready and status ok (default wait 60 seconds)
    pub async fn wait_for_instance_ready(
        &self,
        instance_id: &str,
        duration: Option<Duration>,
    ) -> Result<(), EC2Error> {
        self.client
            .wait_until_instance_status_ok()
            .instance_ids(instance_id)
            .wait(duration.unwrap_or(Duration::from_secs(60)))
            .await
            .map_err(|err| match err {
                WaiterError::ExceededMaxWait(exceeded) => EC2Error(format!(
                    "Exceeded max time ({}s) waiting for instance to start.",
                    exceeded.max_wait().as_secs()
                )),
                _ => EC2Error::from(err),
            })?;
        Ok(())
    }

    pub async fn describe_instance(&self, instance_id: &str) -> Result<Instance, EC2Error> {
        let response = self
            .client
            .describe_instances()
            .instance_ids(instance_id)
            .send()
            .await?;

        let instance = response
            .reservations()
            .first()
            .ok_or_else(|| EC2Error::new(format!("No instance reservations for {instance_id}")))?
            .instances()
            .first()
            .ok_or_else(|| {
                EC2Error::new(format!("No instances in reservation for {instance_id}"))
            })?;

        Ok(instance.clone())
    }

    pub async fn start_instance(&self, instance_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Starting instance {instance_id}");

        self.client
            .start_instances()
            .instance_ids(instance_id)
            .send()
            .await?;

        tracing::info!("Started instance.");

        Ok(())
    }

    pub async fn stop_instance(&self, instance_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Stopping instance {instance_id}");

        self.client
            .stop_instances()
            .instance_ids(instance_id)
            .send()
            .await?;

        self.wait_for_instance_stopped(instance_id, None).await?;

        tracing::info!("Stopped instance.");

        Ok(())
    }

    pub async fn reboot_instance(&self, instance_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Rebooting instance {instance_id}");

        self.client
            .reboot_instances()
            .instance_ids(instance_id)
            .send()
            .await?;

        Ok(())
    }

    pub async fn wait_for_instance_stopped(
        &self,
        instance_id: &str,
        duration: Option<Duration>,
    ) -> Result<(), EC2Error> {
        self.client
            .wait_until_instance_stopped()
            .instance_ids(instance_id)
            .wait(duration.unwrap_or(Duration::from_secs(60)))
            .await
            .map_err(|err| match err {
                WaiterError::ExceededMaxWait(exceeded) => EC2Error(format!(
                    "Exceeded max time ({}s) waiting for instance to stop.",
                    exceeded.max_wait().as_secs(),
                )),
                _ => EC2Error::from(err),
            })?;
        Ok(())
    }

    pub async fn delete_instance(&self, instance_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Deleting instance with id {instance_id}");
        self.stop_instance(instance_id).await?;
        self.client
            .terminate_instances()
            .instance_ids(instance_id)
            .send()
            .await?;
        self.wait_for_instance_terminated(instance_id).await?;
        tracing::info!("Terminated instance with id {instance_id}");
        Ok(())
    }

    async fn wait_for_instance_terminated(&self, instance_id: &str) -> Result<(), EC2Error> {
        self.client
            .wait_until_instance_terminated()
            .instance_ids(instance_id)
            .wait(Duration::from_secs(60))
            .await
            .map_err(|err| match err {
                WaiterError::ExceededMaxWait(exceeded) => EC2Error(format!(
                    "Exceeded max time ({}s) waiting for instance to terminate.",
                    exceeded.max_wait().as_secs(),
                )),
                _ => EC2Error::from(err),
            })?;
        Ok(())
    }

    pub async fn allocate_ip_address(&self) -> Result<AllocateAddressOutput, EC2Error> {
        self.client
            .allocate_address()
            .domain(DomainType::Vpc)
            .send()
            .await
            .map_err(EC2Error::from)
    }

    pub async fn deallocate_ip_address(&self, allocation_id: &str) -> Result<(), EC2Error> {
        self.client
            .release_address()
            .allocation_id(allocation_id)
            .send()
            .await?;
        Ok(())
    }

    pub async fn associate_ip_address(
        &self,
        allocation_id: &str,
        instance_id: &str,
    ) -> Result<AssociateAddressOutput, EC2Error> {
        let response = self
            .client
            .associate_address()
            .allocation_id(allocation_id)
            .instance_id(instance_id)
            .send()
            .await?;
        Ok(response)
    }

    pub async fn disassociate_ip_address(&self, association_id: &str) -> Result<(), EC2Error> {
        self.client
            .disassociate_address()
            .association_id(association_id)
            .send()
            .await?;
        Ok(())
    }
}

#[derive(Debug)]
pub struct EC2Error(String);
impl EC2Error {
    pub fn new(value: impl Into<String>) -> Self {
        EC2Error(value.into())
    }

    pub fn add_message(self, message: impl Into<String>) -> Self {
        EC2Error(format!("{}: {}", message.into(), self.0))
    }
}

impl<T: ProvideErrorMetadata> From<T> for EC2Error {
    fn from(value: T) -> Self {
        EC2Error(format!(
            "{}: {}",
            value
                .code()
                .map(String::from)
                .unwrap_or("unknown code".into()),
            value
                .message()
                .map(String::from)
                .unwrap_or("missing reason".into()),
        ))
    }
}

impl std::error::Error for EC2Error {}

impl std::fmt::Display for EC2Error {
    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
        write!(f, "{}", self.0)
    }
}
```
SSM 구조체는 테스트를 위한 자동 모의점 역할을 하며 해당 함수는 SSM SDK 직접 호출을 래핑합니다.  

```
use aws_sdk_ssm::{types::Parameter, Client};
use aws_smithy_async::future::pagination_stream::TryFlatMap;

use crate::ec2::EC2Error;

#[cfg(test)]
use mockall::automock;

#[cfg(not(test))]
pub use SSMImpl as SSM;

#[cfg(test)]
pub use MockSSMImpl as SSM;

pub struct SSMImpl {
    inner: Client,
}

#[cfg_attr(test, automock)]
impl SSMImpl {
    pub fn new(inner: Client) -> Self {
        SSMImpl { inner }
    }

    pub async fn list_path(&self, path: &str) -> Result<Vec<Parameter>, EC2Error> {
        let maybe_params: Vec<Result<Parameter, _>> = TryFlatMap::new(
            self.inner
                .get_parameters_by_path()
                .path(path)
                .into_paginator()
                .send(),
        )
        .flat_map(|item| item.parameters.unwrap_or_default())
        .collect()
        .await;
        // Fail on the first error
        let params = maybe_params
            .into_iter()
            .collect::<Result<Vec<Parameter>, _>>()?;
        Ok(params)
    }
}
```
시나리오는 여러 '관리자' 스타일 구조체를 사용하여 시나리오 전체에서 생성 및 삭제된 리소스에 대한 액세스를 처리합니다.  

```
use aws_sdk_ec2::operation::{
    allocate_address::AllocateAddressOutput, associate_address::AssociateAddressOutput,
};

use crate::ec2::{EC2Error, EC2};

/// ElasticIpManager tracks the lifecycle of a public IP address, including its
/// allocation from the global pool and association with a specific instance.
#[derive(Debug, Default)]
pub struct ElasticIpManager {
    elastic_ip: Option<AllocateAddressOutput>,
    association: Option<AssociateAddressOutput>,
}

impl ElasticIpManager {
    pub fn has_allocation(&self) -> bool {
        self.elastic_ip.is_some()
    }

    pub fn public_ip(&self) -> &str {
        if let Some(allocation) = &self.elastic_ip {
            if let Some(addr) = allocation.public_ip() {
                return addr;
            }
        }
        "0.0.0.0"
    }

    pub async fn allocate(&mut self, ec2: &EC2) -> Result<(), EC2Error> {
        let allocation = ec2.allocate_ip_address().await?;
        self.elastic_ip = Some(allocation);
        Ok(())
    }

    pub async fn associate(&mut self, ec2: &EC2, instance_id: &str) -> Result<(), EC2Error> {
        if let Some(allocation) = &self.elastic_ip {
            if let Some(allocation_id) = allocation.allocation_id() {
                let association = ec2.associate_ip_address(allocation_id, instance_id).await?;
                self.association = Some(association);
                return Ok(());
            }
        }
        Err(EC2Error::new("No ip address allocation to associate"))
    }

    pub async fn remove(mut self, ec2: &EC2) -> Result<(), EC2Error> {
        if let Some(association) = &self.association {
            if let Some(association_id) = association.association_id() {
                ec2.disassociate_ip_address(association_id).await?;
            }
        }
        self.association = None;
        if let Some(allocation) = &self.elastic_ip {
            if let Some(allocation_id) = allocation.allocation_id() {
                ec2.deallocate_ip_address(allocation_id).await?;
            }
        }
        self.elastic_ip = None;
        Ok(())
    }
}


use std::fmt::Display;

use aws_sdk_ec2::types::{Instance, InstanceType, KeyPairInfo, SecurityGroup};

use crate::ec2::{EC2Error, EC2};

/// InstanceManager wraps the lifecycle of an EC2 Instance.
#[derive(Debug, Default)]
pub struct InstanceManager {
    instance: Option<Instance>,
}

impl InstanceManager {
    pub fn instance_id(&self) -> &str {
        if let Some(instance) = &self.instance {
            if let Some(id) = instance.instance_id() {
                return id;
            }
        }
        "Unknown"
    }

    pub fn instance_name(&self) -> &str {
        if let Some(instance) = &self.instance {
            if let Some(tag) = instance.tags().iter().find(|e| e.key() == Some("Name")) {
                if let Some(value) = tag.value() {
                    return value;
                }
            }
        }
        "Unknown"
    }

    pub fn instance_ip(&self) -> &str {
        if let Some(instance) = &self.instance {
            if let Some(public_ip_address) = instance.public_ip_address() {
                return public_ip_address;
            }
        }
        "0.0.0.0"
    }

    pub fn instance_display_name(&self) -> String {
        format!("{} ({})", self.instance_name(), self.instance_id())
    }

    /// Create an EC2 instance with the given ID on a given type, using a
    /// generated KeyPair and applying a list of security groups.
    pub async fn create(
        &mut self,
        ec2: &EC2,
        image_id: &str,
        instance_type: InstanceType,
        key_pair: &KeyPairInfo,
        security_groups: Vec<&SecurityGroup>,
    ) -> Result<(), EC2Error> {
        let instance_id = ec2
            .create_instance(image_id, instance_type, key_pair, security_groups)
            .await?;
        let instance = ec2.describe_instance(&instance_id).await?;
        self.instance = Some(instance);
        Ok(())
    }

    /// Start the managed EC2 instance, if present.
    pub async fn start(&self, ec2: &EC2) -> Result<(), EC2Error> {
        if self.instance.is_some() {
            ec2.start_instance(self.instance_id()).await?;
        }
        Ok(())
    }

    /// Stop the managed EC2 instance, if present.
    pub async fn stop(&self, ec2: &EC2) -> Result<(), EC2Error> {
        if self.instance.is_some() {
            ec2.stop_instance(self.instance_id()).await?;
        }
        Ok(())
    }

    pub async fn reboot(&self, ec2: &EC2) -> Result<(), EC2Error> {
        if self.instance.is_some() {
            ec2.reboot_instance(self.instance_id()).await?;
            ec2.wait_for_instance_stopped(self.instance_id(), None)
                .await?;
            ec2.wait_for_instance_ready(self.instance_id(), None)
                .await?;
        }
        Ok(())
    }

    /// Terminate and delete the managed EC2 instance, if present.
    pub async fn delete(self, ec2: &EC2) -> Result<(), EC2Error> {
        if self.instance.is_some() {
            ec2.delete_instance(self.instance_id()).await?;
        }
        Ok(())
    }
}

impl Display for InstanceManager {
    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
        if let Some(instance) = &self.instance {
            writeln!(f, "\tID: {}", instance.instance_id().unwrap_or("(Unknown)"))?;
            writeln!(
                f,
                "\tImage ID: {}",
                instance.image_id().unwrap_or("(Unknown)")
            )?;
            writeln!(
                f,
                "\tInstance type: {}",
                instance
                    .instance_type()
                    .map(|it| format!("{it}"))
                    .unwrap_or("(Unknown)".to_string())
            )?;
            writeln!(
                f,
                "\tKey name: {}",
                instance.key_name().unwrap_or("(Unknown)")
            )?;
            writeln!(f, "\tVPC ID: {}", instance.vpc_id().unwrap_or("(Unknown)"))?;
            writeln!(
                f,
                "\tPublic IP: {}",
                instance.public_ip_address().unwrap_or("(Unknown)")
            )?;
            let instance_state = instance
                .state
                .as_ref()
                .map(|is| {
                    is.name()
                        .map(|isn| format!("{isn}"))
                        .unwrap_or("(Unknown)".to_string())
                })
                .unwrap_or("(Unknown)".to_string());
            writeln!(f, "\tState: {instance_state}")?;
        } else {
            writeln!(f, "\tNo loaded instance")?;
        }
        Ok(())
    }
}


use std::{env, path::PathBuf};

use aws_sdk_ec2::types::KeyPairInfo;

use crate::ec2::{EC2Error, EC2};

use super::util::Util;

/// KeyPairManager tracks a KeyPairInfo and the path the private key has been
/// written to, if it's been created.
#[derive(Debug)]
pub struct KeyPairManager {
    key_pair: KeyPairInfo,
    key_file_path: Option<PathBuf>,
    key_file_dir: PathBuf,
}

impl KeyPairManager {
    pub fn new() -> Self {
        Self::default()
    }

    pub fn key_pair(&self) -> &KeyPairInfo {
        &self.key_pair
    }

    pub fn key_file_path(&self) -> Option<&PathBuf> {
        self.key_file_path.as_ref()
    }

    pub fn key_file_dir(&self) -> &PathBuf {
        &self.key_file_dir
    }

    /// Creates a key pair that can be used to securely connect to an EC2 instance.
    /// The returned key pair contains private key information that cannot be retrieved
    /// again. The private key data is stored as a .pem file.
    ///
    /// :param key_name: The name of the key pair to create.
    pub async fn create(
        &mut self,
        ec2: &EC2,
        util: &Util,
        key_name: String,
    ) -> Result<KeyPairInfo, EC2Error> {
        let (key_pair, material) = ec2.create_key_pair(key_name.clone()).await.map_err(|e| {
            self.key_pair = KeyPairInfo::builder().key_name(key_name.clone()).build();
            e.add_message(format!("Couldn't create key {key_name}"))
        })?;

        let path = self.key_file_dir.join(format!("{key_name}.pem"));

        // Save the key_pair information immediately, so it can get cleaned up if write_secure fails.
        self.key_file_path = Some(path.clone());
        self.key_pair = key_pair.clone();

        util.write_secure(&key_name, &path, material)?;

        Ok(key_pair)
    }

    pub async fn delete(self, ec2: &EC2, util: &Util) -> Result<(), EC2Error> {
        if let Some(key_name) = self.key_pair.key_name() {
            ec2.delete_key_pair(key_name).await?;
            if let Some(key_path) = self.key_file_path() {
                if let Err(err) = util.remove(key_path) {
                    eprintln!("Failed to remove {key_path:?} ({err:?})");
                }
            }
        }
        Ok(())
    }

    pub async fn list(&self, ec2: &EC2) -> Result<Vec<KeyPairInfo>, EC2Error> {
        ec2.list_key_pair().await
    }
}

impl Default for KeyPairManager {
    fn default() -> Self {
        KeyPairManager {
            key_pair: KeyPairInfo::builder().build(),
            key_file_path: Default::default(),
            key_file_dir: env::temp_dir(),
        }
    }
}


use std::net::Ipv4Addr;

use aws_sdk_ec2::types::SecurityGroup;

use crate::ec2::{EC2Error, EC2};

/// SecurityGroupManager tracks the lifecycle of a SecurityGroup for an instance,
/// including adding a rule to allow SSH from a public IP address.
#[derive(Debug, Default)]
pub struct SecurityGroupManager {
    group_name: String,
    group_description: String,
    security_group: Option<SecurityGroup>,
}

impl SecurityGroupManager {
    pub async fn create(
        &mut self,
        ec2: &EC2,
        group_name: &str,
        group_description: &str,
    ) -> Result<(), EC2Error> {
        self.group_name = group_name.into();
        self.group_description = group_description.into();

        self.security_group = Some(
            ec2.create_security_group(group_name, group_description)
                .await
                .map_err(|e| e.add_message("Couldn't create security group"))?,
        );

        Ok(())
    }

    pub async fn authorize_ingress(&self, ec2: &EC2, ip_address: Ipv4Addr) -> Result<(), EC2Error> {
        if let Some(sg) = &self.security_group {
            ec2.authorize_security_group_ssh_ingress(
                sg.group_id()
                    .ok_or_else(|| EC2Error::new("Missing security group ID"))?,
                vec![ip_address],
            )
            .await?;
        };

        Ok(())
    }

    pub async fn delete(self, ec2: &EC2) -> Result<(), EC2Error> {
        if let Some(sg) = &self.security_group {
            ec2.delete_security_group(
                sg.group_id()
                    .ok_or_else(|| EC2Error::new("Missing security group ID"))?,
            )
            .await?;
        };

        Ok(())
    }

    pub fn group_name(&self) -> &str {
        &self.group_name
    }

    pub fn vpc_id(&self) -> Option<&str> {
        self.security_group.as_ref().and_then(|sg| sg.vpc_id())
    }

    pub fn security_group(&self) -> Option<&SecurityGroup> {
        self.security_group.as_ref()
    }
}

impl std::fmt::Display for SecurityGroupManager {
    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
        match &self.security_group {
            Some(sg) => {
                writeln!(
                    f,
                    "Security group: {}",
                    sg.group_name().unwrap_or("(unknown group)")
                )?;
                writeln!(f, "\tID: {}", sg.group_id().unwrap_or("(unknown group id)"))?;
                writeln!(f, "\tVPC: {}", sg.vpc_id().unwrap_or("(unknown group vpc)"))?;
                if !sg.ip_permissions().is_empty() {
                    writeln!(f, "\tInbound Permissions:")?;
                    for permission in sg.ip_permissions() {
                        writeln!(f, "\t\t{permission:?}")?;
                    }
                }
                Ok(())
            }
            None => writeln!(f, "No security group loaded."),
        }
    }
}
```
시나리오의 기본 진입점입니다.  

```
use ec2_code_examples::{
    ec2::EC2,
    getting_started::{
        scenario::{run, Ec2InstanceScenario},
        util::UtilImpl,
    },
    ssm::SSM,
};

#[tokio::main]
async fn main() {
    tracing_subscriber::fmt::init();
    let sdk_config = aws_config::load_from_env().await;
    let ec2 = EC2::new(aws_sdk_ec2::Client::new(&sdk_config));
    let ssm = SSM::new(aws_sdk_ssm::Client::new(&sdk_config));
    let util = UtilImpl {};
    let scenario = Ec2InstanceScenario::new(ec2, ssm, util);
    run(scenario).await;
}
```
+ API 세부 정보는 *AWS SDK for Rust API 참조*의 다음 주제를 참조하세요.
  + [AllocateAddress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.allocate_address)
  + [AssociateAddress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.associate_address)
  + [AuthorizeSecurityGroupIngress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.authorize_security_group_ingress)
  + [CreateKeyPair](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.create_key_pair)
  + [CreateSecurityGroup](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.create_security_group)
  + [DeleteKeyPair](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.delete_key_pair)
  + [DeleteSecurityGroup](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.delete_security_group)
  + [DescribeImages](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_images)
  + [DescribeInstanceTypes](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_instance_types)
  + [DescribeInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_instances)
  + [DescribeKeyPairs](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_key_pairs)
  + [DescribeSecurityGroups](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_security_groups)
  + [DisassociateAddress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.disassociate_address)
  + [ReleaseAddress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.release_address)
  + [RunInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.run_instances)
  + [StartInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.start_instances)
  + [StopInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.stop_instances)
  + [TerminateInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.terminate_instances)
  + [UnmonitorInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.unmonitor_instances)

## 작업
<a name="actions"></a>

### `AllocateAddress`
<a name="ec2_AllocateAddress_rust_topic"></a>

다음 코드 예시는 `AllocateAddress`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn allocate_ip_address(&self) -> Result<AllocateAddressOutput, EC2Error> {
        self.client
            .allocate_address()
            .domain(DomainType::Vpc)
            .send()
            .await
            .map_err(EC2Error::from)
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [AllocateAddress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.allocate_address)를 참조하세요.

### `AssociateAddress`
<a name="ec2_AssociateAddress_rust_topic"></a>

다음 코드 예시는 `AssociateAddress`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn associate_ip_address(
        &self,
        allocation_id: &str,
        instance_id: &str,
    ) -> Result<AssociateAddressOutput, EC2Error> {
        let response = self
            .client
            .associate_address()
            .allocation_id(allocation_id)
            .instance_id(instance_id)
            .send()
            .await?;
        Ok(response)
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [AssociateAddress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.associate_address)를 참조하세요.

### `AuthorizeSecurityGroupIngress`
<a name="ec2_AuthorizeSecurityGroupIngress_rust_topic"></a>

다음 코드 예시는 `AuthorizeSecurityGroupIngress`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    /// Add an ingress rule to a security group explicitly allowing IPv4 address
    /// as {ip}/32 over TCP port 22.
    pub async fn authorize_security_group_ssh_ingress(
        &self,
        group_id: &str,
        ingress_ips: Vec<Ipv4Addr>,
    ) -> Result<(), EC2Error> {
        tracing::info!("Authorizing ingress for security group {group_id}");
        self.client
            .authorize_security_group_ingress()
            .group_id(group_id)
            .set_ip_permissions(Some(
                ingress_ips
                    .into_iter()
                    .map(|ip| {
                        IpPermission::builder()
                            .ip_protocol("tcp")
                            .from_port(22)
                            .to_port(22)
                            .ip_ranges(IpRange::builder().cidr_ip(format!("{ip}/32")).build())
                            .build()
                    })
                    .collect(),
            ))
            .send()
            .await?;
        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [AuthorizeSecurityGroupIngress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.authorize_security_group_ingress)를 참조하세요.

### `CreateKeyPair`
<a name="ec2_CreateKeyPair_rust_topic"></a>

다음 코드 예시는 `CreateKeyPair`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.
EC2 클라이언트의 create\$1key\$1pair를 직접적으로 호출하고 반환된 값을 추출하는 Rust 구현입니다.  

```
    pub async fn create_key_pair(&self, name: String) -> Result<(KeyPairInfo, String), EC2Error> {
        tracing::info!("Creating key pair {name}");
        let output = self.client.create_key_pair().key_name(name).send().await?;
        let info = KeyPairInfo::builder()
            .set_key_name(output.key_name)
            .set_key_fingerprint(output.key_fingerprint)
            .set_key_pair_id(output.key_pair_id)
            .build();
        let material = output
            .key_material
            .ok_or_else(|| EC2Error::new("Create Key Pair has no key material"))?;
        Ok((info, material))
    }
```
create\$1key impl을 직접적으로 호출하고 PEM 프라이빗 키를 안전하게 저장하는 함수입니다.  

```
    /// Creates a key pair that can be used to securely connect to an EC2 instance.
    /// The returned key pair contains private key information that cannot be retrieved
    /// again. The private key data is stored as a .pem file.
    ///
    /// :param key_name: The name of the key pair to create.
    pub async fn create(
        &mut self,
        ec2: &EC2,
        util: &Util,
        key_name: String,
    ) -> Result<KeyPairInfo, EC2Error> {
        let (key_pair, material) = ec2.create_key_pair(key_name.clone()).await.map_err(|e| {
            self.key_pair = KeyPairInfo::builder().key_name(key_name.clone()).build();
            e.add_message(format!("Couldn't create key {key_name}"))
        })?;

        let path = self.key_file_dir.join(format!("{key_name}.pem"));

        // Save the key_pair information immediately, so it can get cleaned up if write_secure fails.
        self.key_file_path = Some(path.clone());
        self.key_pair = key_pair.clone();

        util.write_secure(&key_name, &path, material)?;

        Ok(key_pair)
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [CreateKeyPair](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.create_key_pair)를 참조하세요.

### `CreateSecurityGroup`
<a name="ec2_CreateSecurityGroup_rust_topic"></a>

다음 코드 예시는 `CreateSecurityGroup`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn create_security_group(
        &self,
        name: &str,
        description: &str,
    ) -> Result<SecurityGroup, EC2Error> {
        tracing::info!("Creating security group {name}");
        let create_output = self
            .client
            .create_security_group()
            .group_name(name)
            .description(description)
            .send()
            .await
            .map_err(EC2Error::from)?;

        let group_id = create_output
            .group_id
            .ok_or_else(|| EC2Error::new("Missing security group id after creation"))?;

        let group = self
            .describe_security_group(&group_id)
            .await?
            .ok_or_else(|| {
                EC2Error::new(format!("Could not find security group with id {group_id}"))
            })?;

        tracing::info!("Created security group {name} as {group_id}");

        Ok(group)
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [CreateSecurityGroup](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.create_security_group)을 참조하세요.

### `CreateTags`
<a name="ec2_CreateTags_rust_topic"></a>

다음 코드 예시는 `CreateTags`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.
이 예제에서는 인스턴스를 만든 후 이름 태그를 적용합니다.  

```
    pub async fn create_instance<'a>(
        &self,
        image_id: &'a str,
        instance_type: InstanceType,
        key_pair: &'a KeyPairInfo,
        security_groups: Vec<&'a SecurityGroup>,
    ) -> Result<String, EC2Error> {
        let run_instances = self
            .client
            .run_instances()
            .image_id(image_id)
            .instance_type(instance_type)
            .key_name(
                key_pair
                    .key_name()
                    .ok_or_else(|| EC2Error::new("Missing key name when launching instance"))?,
            )
            .set_security_group_ids(Some(
                security_groups
                    .iter()
                    .filter_map(|sg| sg.group_id.clone())
                    .collect(),
            ))
            .min_count(1)
            .max_count(1)
            .send()
            .await?;

        if run_instances.instances().is_empty() {
            return Err(EC2Error::new("Failed to create instance"));
        }

        let instance_id = run_instances.instances()[0].instance_id().unwrap();
        let response = self
            .client
            .create_tags()
            .resources(instance_id)
            .tags(
                Tag::builder()
                    .key("Name")
                    .value("From SDK Examples")
                    .build(),
            )
            .send()
            .await;

        match response {
            Ok(_) => tracing::info!("Created {instance_id} and applied tags."),
            Err(err) => {
                tracing::info!("Error applying tags to {instance_id}: {err:?}");
                return Err(err.into());
            }
        }

        tracing::info!("Instance is created.");

        Ok(instance_id.to_string())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [CreateTags](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.create_tags)를 참조하세요.

### `DeleteKeyPair`
<a name="ec2_DeleteKeyPair_rust_topic"></a>

다음 코드 예시는 `DeleteKeyPair`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.
백업 프라이빗 PEM 키를 함께 제거하는 delete\$1key 주위의 래퍼입니다.  

```
    pub async fn delete(self, ec2: &EC2, util: &Util) -> Result<(), EC2Error> {
        if let Some(key_name) = self.key_pair.key_name() {
            ec2.delete_key_pair(key_name).await?;
            if let Some(key_path) = self.key_file_path() {
                if let Err(err) = util.remove(key_path) {
                    eprintln!("Failed to remove {key_path:?} ({err:?})");
                }
            }
        }
        Ok(())
    }
```

```
    pub async fn delete_key_pair(&self, key_name: &str) -> Result<(), EC2Error> {
        let key_name: String = key_name.into();
        tracing::info!("Deleting key pair {key_name}");
        self.client
            .delete_key_pair()
            .key_name(key_name)
            .send()
            .await?;
        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DeleteKeyPair](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.delete_key_pair)를 참조하세요.

### `DeleteSecurityGroup`
<a name="ec2_DeleteSecurityGroup_rust_topic"></a>

다음 코드 예시는 `DeleteSecurityGroup`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn delete_security_group(&self, group_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Deleting security group {group_id}");
        self.client
            .delete_security_group()
            .group_id(group_id)
            .send()
            .await?;
        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DeleteSecurityGroup](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.delete_security_group)을 참조하세요.

### `DeleteSnapshot`
<a name="ec2_DeleteSnapshot_rust_topic"></a>

다음 코드 예시는 `DeleteSnapshot`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ebs#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
async fn delete_snapshot(client: &Client, id: &str) -> Result<(), Error> {
    client.delete_snapshot().snapshot_id(id).send().await?;

    println!("Deleted");

    Ok(())
}
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DeleteSnapshot](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.delete_snapshot)을 참조하세요.

### `DescribeImages`
<a name="ec2_DescribeImages_rust_topic"></a>

다음 코드 예시는 `DescribeImages`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn list_images(&self, ids: Vec<Parameter>) -> Result<Vec<Image>, EC2Error> {
        let image_ids = ids.into_iter().filter_map(|p| p.value).collect();
        let output = self
            .client
            .describe_images()
            .set_image_ids(Some(image_ids))
            .send()
            .await?;

        let images = output.images.unwrap_or_default();
        if images.is_empty() {
            Err(EC2Error::new("No images for selected AMIs"))
        } else {
            Ok(images)
        }
    }
```
list\$1images 함수를 SSM과 함께 사용하여 환경에 맞게 제한합니다. SSM에 대한 자세한 내용은 https://docs.aws.amazon.com/systems-manager/latest/userguide/example\$1ssm\$1GetParameters\$1section.html 페이지를 참조하세요.  

```
    async fn find_image(&mut self) -> Result<ScenarioImage, EC2Error> {
        let params: Vec<Parameter> = self
            .ssm
            .list_path("/aws/service/ami-amazon-linux-latest")
            .await
            .map_err(|e| e.add_message("Could not find parameters for available images"))?
            .into_iter()
            .filter(|param| param.name().is_some_and(|name| name.contains("amzn2")))
            .collect();
        let amzn2_images: Vec<ScenarioImage> = self
            .ec2
            .list_images(params)
            .await
            .map_err(|e| e.add_message("Could not find images"))?
            .into_iter()
            .map(ScenarioImage::from)
            .collect();
        println!("We will now create an instance from an Amazon Linux 2 AMI");
        let ami = self.util.select_scenario_image(amzn2_images)?;
        Ok(ami)
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DescribeImages](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_images)를 참조하세요.

### `DescribeInstanceStatus`
<a name="ec2_DescribeInstanceStatus_rust_topic"></a>

다음 코드 예시는 `DescribeInstanceStatus`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
async fn show_all_events(client: &Client) -> Result<(), Error> {
    let resp = client.describe_regions().send().await.unwrap();

    for region in resp.regions.unwrap_or_default() {
        let reg: &'static str = Box::leak(Box::from(region.region_name().unwrap()));
        let region_provider = RegionProviderChain::default_provider().or_else(reg);
        let config = aws_config::from_env().region(region_provider).load().await;
        let new_client = Client::new(&config);

        let resp = new_client.describe_instance_status().send().await;

        println!("Instances in region {}:", reg);
        println!();

        for status in resp.unwrap().instance_statuses() {
            println!(
                "  Events scheduled for instance ID: {}",
                status.instance_id().unwrap_or_default()
            );
            for event in status.events() {
                println!("    Event ID:     {}", event.instance_event_id().unwrap());
                println!("    Description:  {}", event.description().unwrap());
                println!("    Event code:   {}", event.code().unwrap().as_ref());
                println!();
            }
        }
    }

    Ok(())
}
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DescribeInstanceStatus](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_instance_status)을 참조하세요.

### `DescribeInstanceTypes`
<a name="ec2_DescribeInstanceTypes_rust_topic"></a>

다음 코드 예시는 `DescribeInstanceTypes`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    /// List instance types that match an image's architecture and are free tier eligible.
    pub async fn list_instance_types(&self, image: &Image) -> Result<Vec<InstanceType>, EC2Error> {
        let architecture = format!(
            "{}",
            image.architecture().ok_or_else(|| EC2Error::new(format!(
                "Image {:?} does not have a listed architecture",
                image.image_id()
            )))?
        );
        let free_tier_eligible_filter = Filter::builder()
            .name("free-tier-eligible")
            .values("false")
            .build();
        let supported_architecture_filter = Filter::builder()
            .name("processor-info.supported-architecture")
            .values(architecture)
            .build();
        let response = self
            .client
            .describe_instance_types()
            .filters(free_tier_eligible_filter)
            .filters(supported_architecture_filter)
            .send()
            .await?;

        Ok(response
            .instance_types
            .unwrap_or_default()
            .into_iter()
            .filter_map(|iti| iti.instance_type)
            .collect())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DescribeInstanceTypes](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_instance_types)를 참조하세요.

### `DescribeInstances`
<a name="ec2_DescribeInstances_rust_topic"></a>

다음 코드 예시는 `DescribeInstances`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.
EC2 인스턴스에 대한 세부 정보를 검색합니다.  

```
    pub async fn describe_instance(&self, instance_id: &str) -> Result<Instance, EC2Error> {
        let response = self
            .client
            .describe_instances()
            .instance_ids(instance_id)
            .send()
            .await?;

        let instance = response
            .reservations()
            .first()
            .ok_or_else(|| EC2Error::new(format!("No instance reservations for {instance_id}")))?
            .instances()
            .first()
            .ok_or_else(|| {
                EC2Error::new(format!("No instances in reservation for {instance_id}"))
            })?;

        Ok(instance.clone())
    }
```
EC2 인스턴스를 만든 후 세부 정보를 검색하고 저장합니다.  

```
    /// Create an EC2 instance with the given ID on a given type, using a
    /// generated KeyPair and applying a list of security groups.
    pub async fn create(
        &mut self,
        ec2: &EC2,
        image_id: &str,
        instance_type: InstanceType,
        key_pair: &KeyPairInfo,
        security_groups: Vec<&SecurityGroup>,
    ) -> Result<(), EC2Error> {
        let instance_id = ec2
            .create_instance(image_id, instance_type, key_pair, security_groups)
            .await?;
        let instance = ec2.describe_instance(&instance_id).await?;
        self.instance = Some(instance);
        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DescribeInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_instances)을 참조하세요.

### `DescribeKeyPairs`
<a name="ec2_DescribeKeyPairs_rust_topic"></a>

다음 코드 예시는 `DescribeKeyPairs`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn list_key_pair(&self) -> Result<Vec<KeyPairInfo>, EC2Error> {
        let output = self.client.describe_key_pairs().send().await?;
        Ok(output.key_pairs.unwrap_or_default())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DescribeKeyPairs](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_key_pairs)를 참조하세요.

### `DescribeRegions`
<a name="ec2_DescribeRegions_rust_topic"></a>

다음 코드 예시는 `DescribeRegions`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
async fn show_regions(client: &Client) -> Result<(), Error> {
    let rsp = client.describe_regions().send().await?;

    println!("Regions:");
    for region in rsp.regions() {
        println!("  {}", region.region_name().unwrap());
    }

    Ok(())
}
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DescribeRegions](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_regions)을 참조하세요.

### `DescribeSecurityGroups`
<a name="ec2_DescribeSecurityGroups_rust_topic"></a>

다음 코드 예시는 `DescribeSecurityGroups`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
async fn show_security_groups(client: &aws_sdk_ec2::Client, group_ids: Vec<String>) {
    let response = client
        .describe_security_groups()
        .set_group_ids(Some(group_ids))
        .send()
        .await;

    match response {
        Ok(output) => {
            for group in output.security_groups() {
                println!(
                    "Found Security Group {} ({}), vpc id {} and description {}",
                    group.group_name().unwrap_or("unknown"),
                    group.group_id().unwrap_or("id-unknown"),
                    group.vpc_id().unwrap_or("vpcid-unknown"),
                    group.description().unwrap_or("(none)")
                );
            }
        }
        Err(err) => {
            let err = err.into_service_error();
            let meta = err.meta();
            let message = meta.message().unwrap_or("unknown");
            let code = meta.code().unwrap_or("unknown");
            eprintln!("Error listing EC2 Security Groups: ({code}) {message}");
        }
    }
}
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DescribeSecurityGroups](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_security_groups)을 참조하세요.

### `DescribeSnapshots`
<a name="ec2_DescribeSnapshots_rust_topic"></a>

다음 코드 예시는 `DescribeSnapshots`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ebs#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.
스냅샷의 상태를 보여 줍니다.  

```
async fn show_state(client: &Client, id: &str) -> Result<(), Error> {
    let resp = client
        .describe_snapshots()
        .filters(Filter::builder().name("snapshot-id").values(id).build())
        .send()
        .await?;

    println!(
        "State: {}",
        resp.snapshots().first().unwrap().state().unwrap().as_ref()
    );

    Ok(())
}
```

```
async fn show_snapshots(client: &Client) -> Result<(), Error> {
    // "self" represents your account ID.
    // You can list the snapshots for any account by replacing
    // "self" with that account ID.
    let resp = client.describe_snapshots().owner_ids("self").send().await?;
    let snapshots = resp.snapshots();
    let length = snapshots.len();

    for snapshot in snapshots {
        println!(
            "ID:          {}",
            snapshot.snapshot_id().unwrap_or_default()
        );
        println!(
            "Description: {}",
            snapshot.description().unwrap_or_default()
        );
        println!("State:       {}", snapshot.state().unwrap().as_ref());
        println!();
    }

    println!();
    println!("Found {} snapshot(s)", length);
    println!();

    Ok(())
}
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DescribeSnapshots](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.describe_snapshots)을 참조하세요.

### `DisassociateAddress`
<a name="ec2_DisassociateAddress_rust_topic"></a>

다음 코드 예시는 `DisassociateAddress`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn disassociate_ip_address(&self, association_id: &str) -> Result<(), EC2Error> {
        self.client
            .disassociate_address()
            .association_id(association_id)
            .send()
            .await?;
        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [DisassociateAddress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.disassociate_address)를 참조하세요.

### `RebootInstances`
<a name="ec2_RebootInstances_rust_topic"></a>

다음 코드 예시는 `RebootInstances`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn reboot(&self, ec2: &EC2) -> Result<(), EC2Error> {
        if self.instance.is_some() {
            ec2.reboot_instance(self.instance_id()).await?;
            ec2.wait_for_instance_stopped(self.instance_id(), None)
                .await?;
            ec2.wait_for_instance_ready(self.instance_id(), None)
                .await?;
        }
        Ok(())
    }
```

```
    pub async fn reboot_instance(&self, instance_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Rebooting instance {instance_id}");

        self.client
            .reboot_instances()
            .instance_ids(instance_id)
            .send()
            .await?;

        Ok(())
    }
```
Waiters API를 사용하여 인스턴스가 중지 및 준비 상태가 될 때까지 기다립니다. Waiters API를 사용하려면 rust 파일에 'use aws\$1sdk\$1ec2::client::Waiters'가 필요합니다.  

```
    /// Wait for an instance to be ready and status ok (default wait 60 seconds)
    pub async fn wait_for_instance_ready(
        &self,
        instance_id: &str,
        duration: Option<Duration>,
    ) -> Result<(), EC2Error> {
        self.client
            .wait_until_instance_status_ok()
            .instance_ids(instance_id)
            .wait(duration.unwrap_or(Duration::from_secs(60)))
            .await
            .map_err(|err| match err {
                WaiterError::ExceededMaxWait(exceeded) => EC2Error(format!(
                    "Exceeded max time ({}s) waiting for instance to start.",
                    exceeded.max_wait().as_secs()
                )),
                _ => EC2Error::from(err),
            })?;
        Ok(())
    }

    pub async fn wait_for_instance_stopped(
        &self,
        instance_id: &str,
        duration: Option<Duration>,
    ) -> Result<(), EC2Error> {
        self.client
            .wait_until_instance_stopped()
            .instance_ids(instance_id)
            .wait(duration.unwrap_or(Duration::from_secs(60)))
            .await
            .map_err(|err| match err {
                WaiterError::ExceededMaxWait(exceeded) => EC2Error(format!(
                    "Exceeded max time ({}s) waiting for instance to stop.",
                    exceeded.max_wait().as_secs(),
                )),
                _ => EC2Error::from(err),
            })?;
        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [RebootInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.reboot_instances)을 참조하세요.

### `ReleaseAddress`
<a name="ec2_ReleaseAddress_rust_topic"></a>

다음 코드 예시는 `ReleaseAddress`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn deallocate_ip_address(&self, allocation_id: &str) -> Result<(), EC2Error> {
        self.client
            .release_address()
            .allocation_id(allocation_id)
            .send()
            .await?;
        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [ReleaseAddress](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.release_address)를 참조하세요.

### `RunInstances`
<a name="ec2_RunInstances_rust_topic"></a>

다음 코드 예시는 `RunInstances`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn create_instance<'a>(
        &self,
        image_id: &'a str,
        instance_type: InstanceType,
        key_pair: &'a KeyPairInfo,
        security_groups: Vec<&'a SecurityGroup>,
    ) -> Result<String, EC2Error> {
        let run_instances = self
            .client
            .run_instances()
            .image_id(image_id)
            .instance_type(instance_type)
            .key_name(
                key_pair
                    .key_name()
                    .ok_or_else(|| EC2Error::new("Missing key name when launching instance"))?,
            )
            .set_security_group_ids(Some(
                security_groups
                    .iter()
                    .filter_map(|sg| sg.group_id.clone())
                    .collect(),
            ))
            .min_count(1)
            .max_count(1)
            .send()
            .await?;

        if run_instances.instances().is_empty() {
            return Err(EC2Error::new("Failed to create instance"));
        }

        let instance_id = run_instances.instances()[0].instance_id().unwrap();
        let response = self
            .client
            .create_tags()
            .resources(instance_id)
            .tags(
                Tag::builder()
                    .key("Name")
                    .value("From SDK Examples")
                    .build(),
            )
            .send()
            .await;

        match response {
            Ok(_) => tracing::info!("Created {instance_id} and applied tags."),
            Err(err) => {
                tracing::info!("Error applying tags to {instance_id}: {err:?}");
                return Err(err.into());
            }
        }

        tracing::info!("Instance is created.");

        Ok(instance_id.to_string())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [RunInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.run_instances)를 참조하세요.

### `StartInstances`
<a name="ec2_StartInstances_rust_topic"></a>

다음 코드 예시는 `StartInstances`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.
인스턴스 ID로 EC2 인스턴스를 시작합니다.  

```
    pub async fn start_instance(&self, instance_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Starting instance {instance_id}");

        self.client
            .start_instances()
            .instance_ids(instance_id)
            .send()
            .await?;

        tracing::info!("Started instance.");

        Ok(())
    }
```
Waiters API를 사용하여 인스턴스가 준비 및 정상 상태가 될 때까지 기다립니다. Waiters API를 사용하려면 rust 파일에 'use aws\$1sdk\$1ec2::client::Waiters'가 필요합니다.  

```
    /// Wait for an instance to be ready and status ok (default wait 60 seconds)
    pub async fn wait_for_instance_ready(
        &self,
        instance_id: &str,
        duration: Option<Duration>,
    ) -> Result<(), EC2Error> {
        self.client
            .wait_until_instance_status_ok()
            .instance_ids(instance_id)
            .wait(duration.unwrap_or(Duration::from_secs(60)))
            .await
            .map_err(|err| match err {
                WaiterError::ExceededMaxWait(exceeded) => EC2Error(format!(
                    "Exceeded max time ({}s) waiting for instance to start.",
                    exceeded.max_wait().as_secs()
                )),
                _ => EC2Error::from(err),
            })?;
        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [StartInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.start_instances)을 참조하세요.

### `StopInstances`
<a name="ec2_StopInstances_rust_topic"></a>

다음 코드 예시는 `StopInstances`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn stop_instance(&self, instance_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Stopping instance {instance_id}");

        self.client
            .stop_instances()
            .instance_ids(instance_id)
            .send()
            .await?;

        self.wait_for_instance_stopped(instance_id, None).await?;

        tracing::info!("Stopped instance.");

        Ok(())
    }
```
Waiters API를 사용하여 인스턴스가 중지 상태가 될 때까지 기다립니다. Waiters API를 사용하려면 rust 파일에 'use aws\$1sdk\$1ec2::client::Waiters'가 필요합니다.  

```
    pub async fn stop_instance(&self, instance_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Stopping instance {instance_id}");

        self.client
            .stop_instances()
            .instance_ids(instance_id)
            .send()
            .await?;

        self.wait_for_instance_stopped(instance_id, None).await?;

        tracing::info!("Stopped instance.");

        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [StopInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.stop_instances)을 참조하세요.

### `TerminateInstances`
<a name="ec2_TerminateInstances_rust_topic"></a>

다음 코드 예시는 `TerminateInstances`의 사용 방법을 보여줍니다.

**SDK for Rust**  
 GitHub에 더 많은 내용이 있습니다. [AWS 코드 예 리포지토리](https://github.com/awsdocs/aws-doc-sdk-examples/tree/main/rustv1/examples/ec2#code-examples)에서 전체 예를 찾고 설정 및 실행하는 방법을 배워보세요.

```
    pub async fn delete_instance(&self, instance_id: &str) -> Result<(), EC2Error> {
        tracing::info!("Deleting instance with id {instance_id}");
        self.stop_instance(instance_id).await?;
        self.client
            .terminate_instances()
            .instance_ids(instance_id)
            .send()
            .await?;
        self.wait_for_instance_terminated(instance_id).await?;
        tracing::info!("Terminated instance with id {instance_id}");
        Ok(())
    }
```
Waiters API를 사용하여 인스턴스가 종료 상태가 될 때까지 기다립니다. Waiters API를 사용하려면 rust 파일에 'use aws\$1sdk\$1ec2::client::Waiters'가 필요합니다.  

```
    async fn wait_for_instance_terminated(&self, instance_id: &str) -> Result<(), EC2Error> {
        self.client
            .wait_until_instance_terminated()
            .instance_ids(instance_id)
            .wait(Duration::from_secs(60))
            .await
            .map_err(|err| match err {
                WaiterError::ExceededMaxWait(exceeded) => EC2Error(format!(
                    "Exceeded max time ({}s) waiting for instance to terminate.",
                    exceeded.max_wait().as_secs(),
                )),
                _ => EC2Error::from(err),
            })?;
        Ok(())
    }
```
+  API 세부 정보는 *AWS SDK for Rust API 참조*의 [TerminateInstances](https://docs.rs/aws-sdk-ec2/latest/aws_sdk_ec2/client/struct.Client.html#method.terminate_instances)를 참조하세요.