terraform/builtin/providers/aws/resource_aws_instance.go

1169 lines
33 KiB
Go
Raw Normal View History

package aws
import (
2014-10-17 18:12:45 +02:00
"bytes"
"crypto/sha1"
"encoding/base64"
"encoding/hex"
"fmt"
"log"
"strings"
2014-07-01 19:10:11 +02:00
"time"
"github.com/aws/aws-sdk-go/aws"
"github.com/aws/aws-sdk-go/aws/awserr"
"github.com/aws/aws-sdk-go/service/ec2"
"github.com/hashicorp/terraform/helper/hashcode"
2014-07-01 19:10:11 +02:00
"github.com/hashicorp/terraform/helper/resource"
"github.com/hashicorp/terraform/helper/schema"
)
func resourceAwsInstance() *schema.Resource {
return &schema.Resource{
Create: resourceAwsInstanceCreate,
Read: resourceAwsInstanceRead,
Update: resourceAwsInstanceUpdate,
Delete: resourceAwsInstanceDelete,
Importer: &schema.ResourceImporter{
State: schema.ImportStatePassthrough,
},
SchemaVersion: 1,
MigrateState: resourceAwsInstanceMigrateState,
Schema: map[string]*schema.Schema{
"ami": &schema.Schema{
Type: schema.TypeString,
Required: true,
ForceNew: true,
},
"associate_public_ip_address": &schema.Schema{
Type: schema.TypeBool,
ForceNew: true,
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
Optional: true,
},
"availability_zone": &schema.Schema{
Type: schema.TypeString,
Optional: true,
Computed: true,
ForceNew: true,
},
"placement_group": &schema.Schema{
Type: schema.TypeString,
Optional: true,
Computed: true,
ForceNew: true,
},
"instance_type": &schema.Schema{
Type: schema.TypeString,
Required: true,
ForceNew: true,
},
"key_name": &schema.Schema{
Type: schema.TypeString,
Optional: true,
ForceNew: true,
Computed: true,
},
"subnet_id": &schema.Schema{
Type: schema.TypeString,
Optional: true,
Computed: true,
ForceNew: true,
},
"private_ip": &schema.Schema{
Type: schema.TypeString,
Optional: true,
ForceNew: true,
Computed: true,
},
"source_dest_check": &schema.Schema{
Type: schema.TypeBool,
Optional: true,
Default: true,
},
"user_data": &schema.Schema{
Type: schema.TypeString,
Optional: true,
ForceNew: true,
StateFunc: func(v interface{}) string {
switch v.(type) {
case string:
hash := sha1.Sum([]byte(v.(string)))
return hex.EncodeToString(hash[:])
default:
return ""
}
},
},
"security_groups": &schema.Schema{
Type: schema.TypeSet,
Optional: true,
Computed: true,
ForceNew: true,
Elem: &schema.Schema{Type: schema.TypeString},
Set: schema.HashString,
},
2015-03-07 07:14:04 +01:00
"vpc_security_group_ids": &schema.Schema{
Type: schema.TypeSet,
Optional: true,
Computed: true,
Elem: &schema.Schema{Type: schema.TypeString},
2016-02-08 00:51:26 +01:00
Set: schema.HashString,
},
"public_dns": &schema.Schema{
Type: schema.TypeString,
Computed: true,
},
"network_interface_id": &schema.Schema{
Type: schema.TypeString,
Computed: true,
},
"public_ip": &schema.Schema{
Type: schema.TypeString,
Computed: true,
},
"instance_state": &schema.Schema{
Type: schema.TypeString,
Computed: true,
},
"private_dns": &schema.Schema{
Type: schema.TypeString,
Computed: true,
},
"ebs_optimized": &schema.Schema{
Type: schema.TypeBool,
Optional: true,
ForceNew: true,
},
"disable_api_termination": &schema.Schema{
Type: schema.TypeBool,
Optional: true,
},
"instance_initiated_shutdown_behavior": &schema.Schema{
Type: schema.TypeString,
Optional: true,
},
"monitoring": &schema.Schema{
Type: schema.TypeBool,
Optional: true,
},
"iam_instance_profile": &schema.Schema{
2014-09-28 20:51:49 +02:00
Type: schema.TypeString,
ForceNew: true,
Optional: true,
},
"tenancy": &schema.Schema{
Type: schema.TypeString,
Optional: true,
Computed: true,
ForceNew: true,
},
"tags": tagsSchema(),
2014-10-17 18:12:45 +02:00
"block_device": &schema.Schema{
Type: schema.TypeMap,
Optional: true,
Removed: "Split out into three sub-types; see Changelog and Docs",
},
"ebs_block_device": &schema.Schema{
2014-10-17 18:12:45 +02:00
Type: schema.TypeSet,
Optional: true,
Computed: true,
2014-10-17 18:12:45 +02:00
Elem: &schema.Resource{
Schema: map[string]*schema.Schema{
"delete_on_termination": &schema.Schema{
Type: schema.TypeBool,
Optional: true,
Default: true,
ForceNew: true,
},
2014-10-17 18:12:45 +02:00
"device_name": &schema.Schema{
Type: schema.TypeString,
Required: true,
ForceNew: true,
},
"encrypted": &schema.Schema{
Type: schema.TypeBool,
Optional: true,
Computed: true,
ForceNew: true,
},
"iops": &schema.Schema{
Type: schema.TypeInt,
2014-10-17 18:12:45 +02:00
Optional: true,
Computed: true,
2014-10-17 18:12:45 +02:00
ForceNew: true,
},
"snapshot_id": &schema.Schema{
2014-10-17 18:12:45 +02:00
Type: schema.TypeString,
Optional: true,
Computed: true,
2014-10-17 18:12:45 +02:00
ForceNew: true,
},
"volume_size": &schema.Schema{
Type: schema.TypeInt,
Optional: true,
Computed: true,
2014-10-17 18:12:45 +02:00
ForceNew: true,
},
"volume_type": &schema.Schema{
Type: schema.TypeString,
2014-10-17 18:12:45 +02:00
Optional: true,
Computed: true,
2014-10-17 18:12:45 +02:00
ForceNew: true,
},
},
},
Set: func(v interface{}) int {
var buf bytes.Buffer
m := v.(map[string]interface{})
buf.WriteString(fmt.Sprintf("%s-", m["device_name"].(string)))
buf.WriteString(fmt.Sprintf("%s-", m["snapshot_id"].(string)))
return hashcode.String(buf.String())
},
},
2014-10-17 18:12:45 +02:00
"ephemeral_block_device": &schema.Schema{
Type: schema.TypeSet,
Optional: true,
Computed: true,
ForceNew: true,
Elem: &schema.Resource{
Schema: map[string]*schema.Schema{
"device_name": &schema.Schema{
Type: schema.TypeString,
Required: true,
2014-10-17 18:12:45 +02:00
},
"virtual_name": &schema.Schema{
Type: schema.TypeString,
Required: true,
},
2014-10-17 18:12:45 +02:00
},
},
Set: func(v interface{}) int {
var buf bytes.Buffer
m := v.(map[string]interface{})
buf.WriteString(fmt.Sprintf("%s-", m["device_name"].(string)))
buf.WriteString(fmt.Sprintf("%s-", m["virtual_name"].(string)))
return hashcode.String(buf.String())
},
2014-10-17 18:12:45 +02:00
},
providers/aws: add root_block_device to aws_instance AWS provides a single `BlockDeviceMapping` to manage three different kinds of block devices: (a) The root volume (b) Ephemeral storage (c) Additional EBS volumes Each of these types has slightly different semantics [1]. (a) The root volume is defined by the AMI; it can only be customized with `volume_size`, `volume_type`, and `delete_on_termination`. (b) Ephemeral storage is made available based on instance type [2]. It's attached automatically if _no_ block device mappings are specified, and must otherwise be defined with block device mapping entries that contain only DeviceName set to a device like "/dev/sdX" and VirtualName set to "ephemeralN". (c) Additional EBS volumes are controlled by mappings that omit `virtual_name` and can specify `volume_size`, `volume_type`, `delete_on_termination`, `snapshot_id`, and `encryption`. After deciding to ignore root block devices to fix #859, we had users with configurations that were attempting to manage the root block device chime in on #913. Terraform does not have the primitives to be able to properly handle a single collection of resources that is partially managed and partially computed, so our strategy here is to break out logical sub-resources for Terraform and hide the BlockDeviceMapping inside the provider implementation. Now (a) is supported by the `root_block_device` sub-resource, and (b) and (c) are still both merged together under `block_device`, though I have yet to see ephemeral block devices working properly. Looking into possibly separating out `ephemeral_block_device` and `ebs_block_device` sub-resources as well, which seem like the logical next step. We'll wait until the next big release for this, though, since it will break backcompat. [1] http://bit.ly/ec2bdmap [2] http://bit.ly/instancestorebytype Fixes #913 Refs #858
2015-02-18 18:45:30 +01:00
"root_block_device": &schema.Schema{
// TODO: This is a set because we don't support singleton
// sub-resources today. We'll enforce that the set only ever has
providers/aws: add root_block_device to aws_instance AWS provides a single `BlockDeviceMapping` to manage three different kinds of block devices: (a) The root volume (b) Ephemeral storage (c) Additional EBS volumes Each of these types has slightly different semantics [1]. (a) The root volume is defined by the AMI; it can only be customized with `volume_size`, `volume_type`, and `delete_on_termination`. (b) Ephemeral storage is made available based on instance type [2]. It's attached automatically if _no_ block device mappings are specified, and must otherwise be defined with block device mapping entries that contain only DeviceName set to a device like "/dev/sdX" and VirtualName set to "ephemeralN". (c) Additional EBS volumes are controlled by mappings that omit `virtual_name` and can specify `volume_size`, `volume_type`, `delete_on_termination`, `snapshot_id`, and `encryption`. After deciding to ignore root block devices to fix #859, we had users with configurations that were attempting to manage the root block device chime in on #913. Terraform does not have the primitives to be able to properly handle a single collection of resources that is partially managed and partially computed, so our strategy here is to break out logical sub-resources for Terraform and hide the BlockDeviceMapping inside the provider implementation. Now (a) is supported by the `root_block_device` sub-resource, and (b) and (c) are still both merged together under `block_device`, though I have yet to see ephemeral block devices working properly. Looking into possibly separating out `ephemeral_block_device` and `ebs_block_device` sub-resources as well, which seem like the logical next step. We'll wait until the next big release for this, though, since it will break backcompat. [1] http://bit.ly/ec2bdmap [2] http://bit.ly/instancestorebytype Fixes #913 Refs #858
2015-02-18 18:45:30 +01:00
// length zero or one below. When TF gains support for
// sub-resources this can be converted.
Type: schema.TypeSet,
providers/aws: add root_block_device to aws_instance AWS provides a single `BlockDeviceMapping` to manage three different kinds of block devices: (a) The root volume (b) Ephemeral storage (c) Additional EBS volumes Each of these types has slightly different semantics [1]. (a) The root volume is defined by the AMI; it can only be customized with `volume_size`, `volume_type`, and `delete_on_termination`. (b) Ephemeral storage is made available based on instance type [2]. It's attached automatically if _no_ block device mappings are specified, and must otherwise be defined with block device mapping entries that contain only DeviceName set to a device like "/dev/sdX" and VirtualName set to "ephemeralN". (c) Additional EBS volumes are controlled by mappings that omit `virtual_name` and can specify `volume_size`, `volume_type`, `delete_on_termination`, `snapshot_id`, and `encryption`. After deciding to ignore root block devices to fix #859, we had users with configurations that were attempting to manage the root block device chime in on #913. Terraform does not have the primitives to be able to properly handle a single collection of resources that is partially managed and partially computed, so our strategy here is to break out logical sub-resources for Terraform and hide the BlockDeviceMapping inside the provider implementation. Now (a) is supported by the `root_block_device` sub-resource, and (b) and (c) are still both merged together under `block_device`, though I have yet to see ephemeral block devices working properly. Looking into possibly separating out `ephemeral_block_device` and `ebs_block_device` sub-resources as well, which seem like the logical next step. We'll wait until the next big release for this, though, since it will break backcompat. [1] http://bit.ly/ec2bdmap [2] http://bit.ly/instancestorebytype Fixes #913 Refs #858
2015-02-18 18:45:30 +01:00
Optional: true,
Computed: true,
Elem: &schema.Resource{
// "You can only modify the volume size, volume type, and Delete on
// Termination flag on the block device mapping entry for the root
// device volume." - bit.ly/ec2bdmap
Schema: map[string]*schema.Schema{
"delete_on_termination": &schema.Schema{
Type: schema.TypeBool,
Optional: true,
Default: true,
ForceNew: true,
},
"iops": &schema.Schema{
providers/aws: add root_block_device to aws_instance AWS provides a single `BlockDeviceMapping` to manage three different kinds of block devices: (a) The root volume (b) Ephemeral storage (c) Additional EBS volumes Each of these types has slightly different semantics [1]. (a) The root volume is defined by the AMI; it can only be customized with `volume_size`, `volume_type`, and `delete_on_termination`. (b) Ephemeral storage is made available based on instance type [2]. It's attached automatically if _no_ block device mappings are specified, and must otherwise be defined with block device mapping entries that contain only DeviceName set to a device like "/dev/sdX" and VirtualName set to "ephemeralN". (c) Additional EBS volumes are controlled by mappings that omit `virtual_name` and can specify `volume_size`, `volume_type`, `delete_on_termination`, `snapshot_id`, and `encryption`. After deciding to ignore root block devices to fix #859, we had users with configurations that were attempting to manage the root block device chime in on #913. Terraform does not have the primitives to be able to properly handle a single collection of resources that is partially managed and partially computed, so our strategy here is to break out logical sub-resources for Terraform and hide the BlockDeviceMapping inside the provider implementation. Now (a) is supported by the `root_block_device` sub-resource, and (b) and (c) are still both merged together under `block_device`, though I have yet to see ephemeral block devices working properly. Looking into possibly separating out `ephemeral_block_device` and `ebs_block_device` sub-resources as well, which seem like the logical next step. We'll wait until the next big release for this, though, since it will break backcompat. [1] http://bit.ly/ec2bdmap [2] http://bit.ly/instancestorebytype Fixes #913 Refs #858
2015-02-18 18:45:30 +01:00
Type: schema.TypeInt,
Optional: true,
Computed: true,
ForceNew: true,
},
"volume_size": &schema.Schema{
Type: schema.TypeInt,
Optional: true,
Computed: true,
ForceNew: true,
},
"volume_type": &schema.Schema{
Type: schema.TypeString,
Optional: true,
Computed: true,
ForceNew: true,
},
},
},
Set: func(v interface{}) int {
// there can be only one root device; no need to hash anything
return 0
},
providers/aws: add root_block_device to aws_instance AWS provides a single `BlockDeviceMapping` to manage three different kinds of block devices: (a) The root volume (b) Ephemeral storage (c) Additional EBS volumes Each of these types has slightly different semantics [1]. (a) The root volume is defined by the AMI; it can only be customized with `volume_size`, `volume_type`, and `delete_on_termination`. (b) Ephemeral storage is made available based on instance type [2]. It's attached automatically if _no_ block device mappings are specified, and must otherwise be defined with block device mapping entries that contain only DeviceName set to a device like "/dev/sdX" and VirtualName set to "ephemeralN". (c) Additional EBS volumes are controlled by mappings that omit `virtual_name` and can specify `volume_size`, `volume_type`, `delete_on_termination`, `snapshot_id`, and `encryption`. After deciding to ignore root block devices to fix #859, we had users with configurations that were attempting to manage the root block device chime in on #913. Terraform does not have the primitives to be able to properly handle a single collection of resources that is partially managed and partially computed, so our strategy here is to break out logical sub-resources for Terraform and hide the BlockDeviceMapping inside the provider implementation. Now (a) is supported by the `root_block_device` sub-resource, and (b) and (c) are still both merged together under `block_device`, though I have yet to see ephemeral block devices working properly. Looking into possibly separating out `ephemeral_block_device` and `ebs_block_device` sub-resources as well, which seem like the logical next step. We'll wait until the next big release for this, though, since it will break backcompat. [1] http://bit.ly/ec2bdmap [2] http://bit.ly/instancestorebytype Fixes #913 Refs #858
2015-02-18 18:45:30 +01:00
},
},
}
}
func resourceAwsInstanceCreate(d *schema.ResourceData, meta interface{}) error {
2015-04-16 22:05:55 +02:00
conn := meta.(*AWSClient).ec2conn
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
instanceOpts, err := buildAwsInstanceOpts(d, meta)
if err != nil {
return err
}
2014-07-15 06:56:37 +02:00
// Build the creation struct
runOpts := &ec2.RunInstancesInput{
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
BlockDeviceMappings: instanceOpts.BlockDeviceMappings,
DisableApiTermination: instanceOpts.DisableAPITermination,
EbsOptimized: instanceOpts.EBSOptimized,
Monitoring: instanceOpts.Monitoring,
IamInstanceProfile: instanceOpts.IAMInstanceProfile,
ImageId: instanceOpts.ImageID,
InstanceInitiatedShutdownBehavior: instanceOpts.InstanceInitiatedShutdownBehavior,
InstanceType: instanceOpts.InstanceType,
KeyName: instanceOpts.KeyName,
MaxCount: aws.Int64(int64(1)),
MinCount: aws.Int64(int64(1)),
NetworkInterfaces: instanceOpts.NetworkInterfaces,
Placement: instanceOpts.Placement,
PrivateIpAddress: instanceOpts.PrivateIPAddress,
SecurityGroupIds: instanceOpts.SecurityGroupIDs,
SecurityGroups: instanceOpts.SecurityGroups,
SubnetId: instanceOpts.SubnetID,
UserData: instanceOpts.UserData64,
}
2014-07-15 06:56:37 +02:00
// Create the instance
log.Printf("[DEBUG] Run configuration: %s", runOpts)
var runResp *ec2.Reservation
err = resource.Retry(15*time.Second, func() *resource.RetryError {
var err error
runResp, err = conn.RunInstances(runOpts)
// IAM instance profiles can take ~10 seconds to propagate in AWS:
// http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/iam-roles-for-amazon-ec2.html#launch-instance-with-role-console
if isAWSErr(err, "InvalidParameterValue", "Invalid IAM Instance Profile") {
log.Printf("[DEBUG] Invalid IAM Instance Profile referenced, retrying...")
return resource.RetryableError(err)
}
// IAM roles can also take time to propagate in AWS:
if isAWSErr(err, "InvalidParameterValue", " has no associated IAM Roles") {
log.Printf("[DEBUG] IAM Instance Profile appears to have no IAM roles, retrying...")
return resource.RetryableError(err)
}
return resource.NonRetryableError(err)
})
// Warn if the AWS Error involves group ids, to help identify situation
// where a user uses group ids in security_groups for the Default VPC.
// See https://github.com/hashicorp/terraform/issues/3798
if isAWSErr(err, "InvalidParameterValue", "groupId is invalid") {
return fmt.Errorf("Error launching instance, possible mismatch of Security Group IDs and Names. See AWS Instance docs here: %s.\n\n\tAWS Error: %s", "https://terraform.io/docs/providers/aws/r/instance.html", err.(awserr.Error).Message())
}
if err != nil {
return fmt.Errorf("Error launching source instance: %s", err)
}
if runResp == nil || len(runResp.Instances) == 0 {
return fmt.Errorf("Error launching source instance: no instances returned in response")
}
instance := runResp.Instances[0]
log.Printf("[INFO] Instance ID: %s", *instance.InstanceId)
// Store the resulting ID so we can look this up later
d.SetId(*instance.InstanceId)
// Wait for the instance to become running so we can get some attributes
// that aren't available until later.
log.Printf(
"[DEBUG] Waiting for instance (%s) to become running",
*instance.InstanceId)
2014-07-01 19:10:11 +02:00
stateConf := &resource.StateChangeConf{
2014-07-28 18:10:28 +02:00
Pending: []string{"pending"},
Target: []string{"running"},
Refresh: InstanceStateRefreshFunc(conn, *instance.InstanceId),
2014-07-28 18:10:28 +02:00
Timeout: 10 * time.Minute,
Delay: 10 * time.Second,
MinTimeout: 3 * time.Second,
2014-07-01 19:10:11 +02:00
}
instanceRaw, err := stateConf.WaitForState()
if err != nil {
return fmt.Errorf(
"Error waiting for instance (%s) to become ready: %s",
*instance.InstanceId, err)
}
2014-07-01 19:10:11 +02:00
instance = instanceRaw.(*ec2.Instance)
2014-07-15 02:24:10 +02:00
// Initialize the connection info
if instance.PublicIpAddress != nil {
d.SetConnInfo(map[string]string{
"type": "ssh",
"host": *instance.PublicIpAddress,
})
} else if instance.PrivateIpAddress != nil {
d.SetConnInfo(map[string]string{
"type": "ssh",
"host": *instance.PrivateIpAddress,
})
}
2014-07-15 02:24:10 +02:00
// Update if we need to
return resourceAwsInstanceUpdate(d, meta)
}
func resourceAwsInstanceRead(d *schema.ResourceData, meta interface{}) error {
2015-04-16 22:05:55 +02:00
conn := meta.(*AWSClient).ec2conn
resp, err := conn.DescribeInstances(&ec2.DescribeInstancesInput{
InstanceIds: []*string{aws.String(d.Id())},
})
if err != nil {
// If the instance was not found, return nil so that we can show
// that the instance is gone.
if ec2err, ok := err.(awserr.Error); ok && ec2err.Code() == "InvalidInstanceID.NotFound" {
d.SetId("")
return nil
}
// Some other error, report it
return err
}
// If nothing was found, then return no state
if len(resp.Reservations) == 0 {
d.SetId("")
return nil
}
instance := resp.Reservations[0].Instances[0]
if instance.State != nil {
// If the instance is terminated, then it is gone
if *instance.State.Name == "terminated" {
d.SetId("")
return nil
}
d.Set("instance_state", instance.State.Name)
}
if instance.Placement != nil {
d.Set("availability_zone", instance.Placement.AvailabilityZone)
}
if instance.Placement.Tenancy != nil {
d.Set("tenancy", instance.Placement.Tenancy)
}
d.Set("ami", instance.ImageId)
d.Set("instance_type", instance.InstanceType)
d.Set("key_name", instance.KeyName)
d.Set("public_dns", instance.PublicDnsName)
d.Set("public_ip", instance.PublicIpAddress)
d.Set("private_dns", instance.PrivateDnsName)
d.Set("private_ip", instance.PrivateIpAddress)
d.Set("iam_instance_profile", iamInstanceProfileArnToName(instance.IamInstanceProfile))
if len(instance.NetworkInterfaces) > 0 {
for _, ni := range instance.NetworkInterfaces {
if *ni.Attachment.DeviceIndex == 0 {
d.Set("subnet_id", ni.SubnetId)
d.Set("network_interface_id", ni.NetworkInterfaceId)
}
}
} else {
d.Set("subnet_id", instance.SubnetId)
d.Set("network_interface_id", "")
}
d.Set("ebs_optimized", instance.EbsOptimized)
if instance.SubnetId != nil && *instance.SubnetId != "" {
d.Set("source_dest_check", instance.SourceDestCheck)
}
if instance.Monitoring != nil && instance.Monitoring.State != nil {
2015-06-25 18:07:11 +02:00
monitoringState := *instance.Monitoring.State
d.Set("monitoring", monitoringState == "enabled" || monitoringState == "pending")
}
d.Set("tags", tagsToMap(instance.Tags))
// Determine whether we're referring to security groups with
// IDs or names. We use a heuristic to figure this out. By default,
// we use IDs if we're in a VPC. However, if we previously had an
// all-name list of security groups, we use names. Or, if we had any
// IDs, we use IDs.
useID := instance.SubnetId != nil && *instance.SubnetId != ""
if v := d.Get("security_groups"); v != nil {
match := useID
sgs := v.(*schema.Set).List()
if len(sgs) > 0 {
match = false
for _, v := range v.(*schema.Set).List() {
if strings.HasPrefix(v.(string), "sg-") {
match = true
break
}
}
}
useID = match
}
2014-07-15 06:56:37 +02:00
// Build up the security groups
sgs := make([]string, 0, len(instance.SecurityGroups))
if useID {
for _, sg := range instance.SecurityGroups {
sgs = append(sgs, *sg.GroupId)
}
log.Printf("[DEBUG] Setting Security Group IDs: %#v", sgs)
if err := d.Set("vpc_security_group_ids", sgs); err != nil {
return err
2015-03-07 07:16:59 +01:00
}
2016-04-22 18:37:27 +02:00
if err := d.Set("security_groups", []string{}); err != nil {
return err
}
} else {
for _, sg := range instance.SecurityGroups {
sgs = append(sgs, *sg.GroupName)
2014-07-15 06:56:37 +02:00
}
log.Printf("[DEBUG] Setting Security Group Names: %#v", sgs)
if err := d.Set("security_groups", sgs); err != nil {
return err
}
2016-04-22 18:37:27 +02:00
if err := d.Set("vpc_security_group_ids", []string{}); err != nil {
return err
}
2014-07-15 06:56:37 +02:00
}
if err := readBlockDevices(d, instance, conn); err != nil {
2014-10-17 18:12:45 +02:00
return err
}
2016-04-22 18:37:27 +02:00
if _, ok := d.GetOk("ephemeral_block_device"); !ok {
d.Set("ephemeral_block_device", []interface{}{})
}
// Instance attributes
{
attr, err := conn.DescribeInstanceAttribute(&ec2.DescribeInstanceAttributeInput{
Attribute: aws.String("disableApiTermination"),
InstanceId: aws.String(d.Id()),
})
if err != nil {
return err
}
d.Set("disable_api_termination", attr.DisableApiTermination.Value)
}
2014-10-17 18:12:45 +02:00
return nil
}
2014-07-01 19:10:11 +02:00
func resourceAwsInstanceUpdate(d *schema.ResourceData, meta interface{}) error {
2015-04-16 22:05:55 +02:00
conn := meta.(*AWSClient).ec2conn
d.Partial(true)
if err := setTags(conn, d); err != nil {
return err
} else {
d.SetPartial("tags")
}
if d.HasChange("source_dest_check") || d.IsNewResource() {
// SourceDestCheck can only be set on VPC instances // AWS will return an error of InvalidParameterCombination if we attempt
// to modify the source_dest_check of an instance in EC2 Classic
log.Printf("[INFO] Modifying `source_dest_check` on Instance %s", d.Id())
_, err := conn.ModifyInstanceAttribute(&ec2.ModifyInstanceAttributeInput{
InstanceId: aws.String(d.Id()),
SourceDestCheck: &ec2.AttributeBooleanValue{
Value: aws.Bool(d.Get("source_dest_check").(bool)),
},
})
if err != nil {
if ec2err, ok := err.(awserr.Error); ok {
// Toloerate InvalidParameterCombination error in Classic, otherwise
// return the error
if "InvalidParameterCombination" != ec2err.Code() {
return err
}
log.Printf("[WARN] Attempted to modify SourceDestCheck on non VPC instance: %s", ec2err.Message())
}
}
}
if d.HasChange("vpc_security_group_ids") {
var groups []*string
if v := d.Get("vpc_security_group_ids").(*schema.Set); v.Len() > 0 {
for _, v := range v.List() {
groups = append(groups, aws.String(v.(string)))
}
}
_, err := conn.ModifyInstanceAttribute(&ec2.ModifyInstanceAttributeInput{
InstanceId: aws.String(d.Id()),
Groups: groups,
})
if err != nil {
return err
}
}
if d.HasChange("disable_api_termination") {
_, err := conn.ModifyInstanceAttribute(&ec2.ModifyInstanceAttributeInput{
InstanceId: aws.String(d.Id()),
DisableApiTermination: &ec2.AttributeBooleanValue{
Value: aws.Bool(d.Get("disable_api_termination").(bool)),
},
})
if err != nil {
return err
}
}
if d.HasChange("instance_initiated_shutdown_behavior") {
log.Printf("[INFO] Modifying instance %s", d.Id())
_, err := conn.ModifyInstanceAttribute(&ec2.ModifyInstanceAttributeInput{
InstanceId: aws.String(d.Id()),
InstanceInitiatedShutdownBehavior: &ec2.AttributeValue{
Value: aws.String(d.Get("instance_initiated_shutdown_behavior").(string)),
},
})
if err != nil {
return err
}
}
if d.HasChange("monitoring") {
var mErr error
if d.Get("monitoring").(bool) {
log.Printf("[DEBUG] Enabling monitoring for Instance (%s)", d.Id())
_, mErr = conn.MonitorInstances(&ec2.MonitorInstancesInput{
InstanceIds: []*string{aws.String(d.Id())},
})
} else {
log.Printf("[DEBUG] Disabling monitoring for Instance (%s)", d.Id())
_, mErr = conn.UnmonitorInstances(&ec2.UnmonitorInstancesInput{
InstanceIds: []*string{aws.String(d.Id())},
})
}
if mErr != nil {
return fmt.Errorf("[WARN] Error updating Instance monitoring: %s", mErr)
}
}
// TODO(mitchellh): wait for the attributes we modified to
// persist the change...
d.Partial(false)
return resourceAwsInstanceRead(d, meta)
}
func resourceAwsInstanceDelete(d *schema.ResourceData, meta interface{}) error {
2015-04-16 22:05:55 +02:00
conn := meta.(*AWSClient).ec2conn
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
if err := awsTerminateInstance(conn, d.Id()); err != nil {
return err
}
d.SetId("")
return nil
}
2014-07-01 19:10:11 +02:00
// InstanceStateRefreshFunc returns a resource.StateRefreshFunc that is used to watch
// an EC2 instance.
func InstanceStateRefreshFunc(conn *ec2.EC2, instanceID string) resource.StateRefreshFunc {
return func() (interface{}, string, error) {
resp, err := conn.DescribeInstances(&ec2.DescribeInstancesInput{
InstanceIds: []*string{aws.String(instanceID)},
})
2014-07-01 19:10:11 +02:00
if err != nil {
if ec2err, ok := err.(awserr.Error); ok && ec2err.Code() == "InvalidInstanceID.NotFound" {
2014-07-01 19:10:11 +02:00
// Set this to nil as if we didn't find anything.
resp = nil
} else {
log.Printf("Error on InstanceStateRefresh: %s", err)
return nil, "", err
}
}
if resp == nil || len(resp.Reservations) == 0 || len(resp.Reservations[0].Instances) == 0 {
// Sometimes AWS just has consistency issues and doesn't see
// our instance yet. Return an empty state.
return nil, "", nil
}
i := resp.Reservations[0].Instances[0]
return i, *i.State.Name, nil
2014-07-01 19:10:11 +02:00
}
}
2014-10-17 18:12:45 +02:00
func readBlockDevices(d *schema.ResourceData, instance *ec2.Instance, conn *ec2.EC2) error {
ibds, err := readBlockDevicesFromInstance(instance, conn)
if err != nil {
return err
}
if err := d.Set("ebs_block_device", ibds["ebs"]); err != nil {
return err
}
2016-04-22 18:37:27 +02:00
// This handles the import case which needs to be defaulted to empty
if _, ok := d.GetOk("root_block_device"); !ok {
if err := d.Set("root_block_device", []interface{}{}); err != nil {
return err
}
}
if ibds["root"] != nil {
2016-04-22 18:37:27 +02:00
roots := []interface{}{ibds["root"]}
if err := d.Set("root_block_device", roots); err != nil {
return err
}
2014-07-01 19:10:11 +02:00
}
return nil
2014-07-01 19:10:11 +02:00
}
2014-10-17 18:12:45 +02:00
func readBlockDevicesFromInstance(instance *ec2.Instance, conn *ec2.EC2) (map[string]interface{}, error) {
blockDevices := make(map[string]interface{})
blockDevices["ebs"] = make([]map[string]interface{}, 0)
blockDevices["root"] = nil
instanceBlockDevices := make(map[string]*ec2.InstanceBlockDeviceMapping)
for _, bd := range instance.BlockDeviceMappings {
if bd.Ebs != nil {
2015-10-08 14:48:04 +02:00
instanceBlockDevices[*bd.Ebs.VolumeId] = bd
}
}
if len(instanceBlockDevices) == 0 {
return nil, nil
}
volIDs := make([]*string, 0, len(instanceBlockDevices))
for volID := range instanceBlockDevices {
volIDs = append(volIDs, aws.String(volID))
}
// Need to call DescribeVolumes to get volume_size and volume_type for each
// EBS block device
volResp, err := conn.DescribeVolumes(&ec2.DescribeVolumesInput{
VolumeIds: volIDs,
})
if err != nil {
return nil, err
}
for _, vol := range volResp.Volumes {
instanceBd := instanceBlockDevices[*vol.VolumeId]
bd := make(map[string]interface{})
if instanceBd.Ebs != nil && instanceBd.Ebs.DeleteOnTermination != nil {
bd["delete_on_termination"] = *instanceBd.Ebs.DeleteOnTermination
}
if vol.Size != nil {
bd["volume_size"] = *vol.Size
}
if vol.VolumeType != nil {
bd["volume_type"] = *vol.VolumeType
}
if vol.Iops != nil {
bd["iops"] = *vol.Iops
}
if blockDeviceIsRoot(instanceBd, instance) {
blockDevices["root"] = bd
} else {
if instanceBd.DeviceName != nil {
bd["device_name"] = *instanceBd.DeviceName
}
if vol.Encrypted != nil {
bd["encrypted"] = *vol.Encrypted
}
if vol.SnapshotId != nil {
bd["snapshot_id"] = *vol.SnapshotId
}
blockDevices["ebs"] = append(blockDevices["ebs"].([]map[string]interface{}), bd)
}
}
return blockDevices, nil
}
func blockDeviceIsRoot(bd *ec2.InstanceBlockDeviceMapping, instance *ec2.Instance) bool {
2015-10-08 14:48:04 +02:00
return bd.DeviceName != nil &&
instance.RootDeviceName != nil &&
2015-10-08 14:48:04 +02:00
*bd.DeviceName == *instance.RootDeviceName
2014-10-17 18:12:45 +02:00
}
func fetchRootDeviceName(ami string, conn *ec2.EC2) (*string, error) {
if ami == "" {
return nil, fmt.Errorf("Cannot fetch root device name for blank AMI ID.")
}
log.Printf("[DEBUG] Describing AMI %q to get root block device name", ami)
res, err := conn.DescribeImages(&ec2.DescribeImagesInput{
ImageIds: []*string{aws.String(ami)},
})
if err != nil {
return nil, err
}
// For a bad image, we just return nil so we don't block a refresh
if len(res.Images) == 0 {
return nil, nil
}
image := res.Images[0]
rootDeviceName := image.RootDeviceName
// Some AMIs have a RootDeviceName like "/dev/sda1" that does not appear as a
// DeviceName in the BlockDeviceMapping list (which will instead have
// something like "/dev/sda")
//
// While this seems like it breaks an invariant of AMIs, it ends up working
// on the AWS side, and AMIs like this are common enough that we need to
// special case it so Terraform does the right thing.
//
// Our heuristic is: if the RootDeviceName does not appear in the
// BlockDeviceMapping, assume that the DeviceName of the first
// BlockDeviceMapping entry serves as the root device.
rootDeviceNameInMapping := false
for _, bdm := range image.BlockDeviceMappings {
if bdm.DeviceName == image.RootDeviceName {
rootDeviceNameInMapping = true
}
}
if !rootDeviceNameInMapping && len(image.BlockDeviceMappings) > 0 {
rootDeviceName = image.BlockDeviceMappings[0].DeviceName
}
if rootDeviceName == nil {
return nil, fmt.Errorf("[WARN] Error finding Root Device Name for AMI (%s)", ami)
}
return rootDeviceName, nil
2014-10-17 18:12:45 +02:00
}
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
func readBlockDeviceMappingsFromConfig(
d *schema.ResourceData, conn *ec2.EC2) ([]*ec2.BlockDeviceMapping, error) {
blockDevices := make([]*ec2.BlockDeviceMapping, 0)
if v, ok := d.GetOk("ebs_block_device"); ok {
vL := v.(*schema.Set).List()
for _, v := range vL {
bd := v.(map[string]interface{})
ebs := &ec2.EbsBlockDevice{
DeleteOnTermination: aws.Bool(bd["delete_on_termination"].(bool)),
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
if v, ok := bd["snapshot_id"].(string); ok && v != "" {
ebs.SnapshotId = aws.String(v)
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
if v, ok := bd["encrypted"].(bool); ok && v {
ebs.Encrypted = aws.Bool(v)
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
if v, ok := bd["volume_size"].(int); ok && v != 0 {
ebs.VolumeSize = aws.Int64(int64(v))
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
if v, ok := bd["volume_type"].(string); ok && v != "" {
ebs.VolumeType = aws.String(v)
}
if v, ok := bd["iops"].(int); ok && v > 0 {
ebs.Iops = aws.Int64(int64(v))
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
blockDevices = append(blockDevices, &ec2.BlockDeviceMapping{
DeviceName: aws.String(bd["device_name"].(string)),
Ebs: ebs,
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
})
}
}
if v, ok := d.GetOk("ephemeral_block_device"); ok {
vL := v.(*schema.Set).List()
for _, v := range vL {
bd := v.(map[string]interface{})
blockDevices = append(blockDevices, &ec2.BlockDeviceMapping{
DeviceName: aws.String(bd["device_name"].(string)),
VirtualName: aws.String(bd["virtual_name"].(string)),
})
}
}
if v, ok := d.GetOk("root_block_device"); ok {
vL := v.(*schema.Set).List()
if len(vL) > 1 {
return nil, fmt.Errorf("Cannot specify more than one root_block_device.")
}
for _, v := range vL {
bd := v.(map[string]interface{})
ebs := &ec2.EbsBlockDevice{
DeleteOnTermination: aws.Bool(bd["delete_on_termination"].(bool)),
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
if v, ok := bd["volume_size"].(int); ok && v != 0 {
ebs.VolumeSize = aws.Int64(int64(v))
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
if v, ok := bd["volume_type"].(string); ok && v != "" {
ebs.VolumeType = aws.String(v)
}
if v, ok := bd["iops"].(int); ok && v > 0 && *ebs.VolumeType == "io1" {
// Only set the iops attribute if the volume type is io1. Setting otherwise
// can trigger a refresh/plan loop based on the computed value that is given
// from AWS, and prevent us from specifying 0 as a valid iops.
// See https://github.com/hashicorp/terraform/pull/4146
// See https://github.com/hashicorp/terraform/issues/7765
ebs.Iops = aws.Int64(int64(v))
} else if v, ok := bd["iops"].(int); ok && v > 0 && *ebs.VolumeType != "io1" {
// Message user about incompatibility
log.Printf("[WARN] IOPs is only valid for storate type io1 for EBS Volumes")
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
if dn, err := fetchRootDeviceName(d.Get("ami").(string), conn); err == nil {
if dn == nil {
return nil, fmt.Errorf(
"Expected 1 AMI for ID: %s, got none",
d.Get("ami").(string))
}
blockDevices = append(blockDevices, &ec2.BlockDeviceMapping{
DeviceName: dn,
Ebs: ebs,
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
})
} else {
return nil, err
}
}
}
return blockDevices, nil
}
type awsInstanceOpts struct {
BlockDeviceMappings []*ec2.BlockDeviceMapping
DisableAPITermination *bool
EBSOptimized *bool
Monitoring *ec2.RunInstancesMonitoringEnabled
IAMInstanceProfile *ec2.IamInstanceProfileSpecification
ImageID *string
InstanceInitiatedShutdownBehavior *string
InstanceType *string
KeyName *string
NetworkInterfaces []*ec2.InstanceNetworkInterfaceSpecification
Placement *ec2.Placement
PrivateIPAddress *string
SecurityGroupIDs []*string
SecurityGroups []*string
SpotPlacement *ec2.SpotPlacement
SubnetID *string
UserData64 *string
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
func buildAwsInstanceOpts(
d *schema.ResourceData, meta interface{}) (*awsInstanceOpts, error) {
conn := meta.(*AWSClient).ec2conn
opts := &awsInstanceOpts{
DisableAPITermination: aws.Bool(d.Get("disable_api_termination").(bool)),
EBSOptimized: aws.Bool(d.Get("ebs_optimized").(bool)),
ImageID: aws.String(d.Get("ami").(string)),
InstanceType: aws.String(d.Get("instance_type").(string)),
}
if v := d.Get("instance_initiated_shutdown_behavior").(string); v != "" {
opts.InstanceInitiatedShutdownBehavior = aws.String(v)
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
opts.Monitoring = &ec2.RunInstancesMonitoringEnabled{
Enabled: aws.Bool(d.Get("monitoring").(bool)),
}
opts.IAMInstanceProfile = &ec2.IamInstanceProfileSpecification{
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
Name: aws.String(d.Get("iam_instance_profile").(string)),
}
user_data := d.Get("user_data").(string)
// Check whether the user_data is already Base64 encoded; don't double-encode
_, base64DecodeError := base64.StdEncoding.DecodeString(user_data)
if base64DecodeError == nil {
opts.UserData64 = aws.String(user_data)
} else {
opts.UserData64 = aws.String(base64.StdEncoding.EncodeToString([]byte(user_data)))
}
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
// check for non-default Subnet, and cast it to a String
subnet, hasSubnet := d.GetOk("subnet_id")
subnetID := subnet.(string)
// Placement is used for aws_instance; SpotPlacement is used for
// aws_spot_instance_request. They represent the same data. :-|
opts.Placement = &ec2.Placement{
AvailabilityZone: aws.String(d.Get("availability_zone").(string)),
GroupName: aws.String(d.Get("placement_group").(string)),
}
opts.SpotPlacement = &ec2.SpotPlacement{
AvailabilityZone: aws.String(d.Get("availability_zone").(string)),
GroupName: aws.String(d.Get("placement_group").(string)),
}
if v := d.Get("tenancy").(string); v != "" {
opts.Placement.Tenancy = aws.String(v)
}
associatePublicIPAddress := d.Get("associate_public_ip_address").(bool)
var groups []*string
if v := d.Get("security_groups"); v != nil {
// Security group names.
// For a nondefault VPC, you must use security group IDs instead.
// See http://docs.aws.amazon.com/AWSEC2/latest/APIReference/API_RunInstances.html
sgs := v.(*schema.Set).List()
if len(sgs) > 0 && hasSubnet {
log.Printf("[WARN] Deprecated. Attempting to use 'security_groups' within a VPC instance. Use 'vpc_security_group_ids' instead.")
}
for _, v := range sgs {
str := v.(string)
groups = append(groups, aws.String(str))
}
}
if hasSubnet && associatePublicIPAddress {
// If we have a non-default VPC / Subnet specified, we can flag
// AssociatePublicIpAddress to get a Public IP assigned. By default these are not provided.
// You cannot specify both SubnetId and the NetworkInterface.0.* parameters though, otherwise
// you get: Network interfaces and an instance-level subnet ID may not be specified on the same request
// You also need to attach Security Groups to the NetworkInterface instead of the instance,
// to avoid: Network interfaces and an instance-level security groups may not be specified on
// the same request
ni := &ec2.InstanceNetworkInterfaceSpecification{
AssociatePublicIpAddress: aws.Bool(associatePublicIPAddress),
DeviceIndex: aws.Int64(int64(0)),
SubnetId: aws.String(subnetID),
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
Groups: groups,
}
if v, ok := d.GetOk("private_ip"); ok {
ni.PrivateIpAddress = aws.String(v.(string))
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
if v := d.Get("vpc_security_group_ids").(*schema.Set); v.Len() > 0 {
for _, v := range v.List() {
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
ni.Groups = append(ni.Groups, aws.String(v.(string)))
}
}
opts.NetworkInterfaces = []*ec2.InstanceNetworkInterfaceSpecification{ni}
} else {
if subnetID != "" {
opts.SubnetID = aws.String(subnetID)
}
if v, ok := d.GetOk("private_ip"); ok {
opts.PrivateIPAddress = aws.String(v.(string))
}
if opts.SubnetID != nil &&
*opts.SubnetID != "" {
opts.SecurityGroupIDs = groups
} else {
opts.SecurityGroups = groups
}
if v := d.Get("vpc_security_group_ids").(*schema.Set); v.Len() > 0 {
for _, v := range v.List() {
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
opts.SecurityGroupIDs = append(opts.SecurityGroupIDs, aws.String(v.(string)))
}
}
}
if v, ok := d.GetOk("key_name"); ok {
opts.KeyName = aws.String(v.(string))
}
blockDevices, err := readBlockDeviceMappingsFromConfig(d, conn)
if err != nil {
return nil, err
}
if len(blockDevices) > 0 {
opts.BlockDeviceMappings = blockDevices
}
return opts, nil
}
func awsTerminateInstance(conn *ec2.EC2, id string) error {
log.Printf("[INFO] Terminating instance: %s", id)
req := &ec2.TerminateInstancesInput{
InstanceIds: []*string{aws.String(id)},
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
}
if _, err := conn.TerminateInstances(req); err != nil {
return fmt.Errorf("Error terminating instance: %s", err)
}
log.Printf("[DEBUG] Waiting for instance (%s) to become terminated", id)
stateConf := &resource.StateChangeConf{
Pending: []string{"pending", "running", "shutting-down", "stopped", "stopping"},
Target: []string{"terminated"},
provider/aws: spot_instance_request This is an iteration on the great work done by @dalehamel in PRs #2095 and #2109. The core team went back and forth on how to best model Spot Instance Requests, requesting and then rejecting a separate-resource implementation in #2109. After more internal discussion, we landed once again on a separate resource to model Spot Instance Requests. Out of respect for @dalehamel's already-significant donated time, with this I'm attempting to pick up the work to take this across the finish line. Important architectural decisions represented here: * Spot Instance Requests are always of type "persistent", to properly match Terraform's declarative model. * The spot_instance_request resource exports several attributes that are expected to be constantly changing as the spot market changes: spot_bid_status, spot_request_state, and instance_id. Creating additional resource dependencies based on these attributes is not recommended, as Terraform diffs will be continually generated to keep up with the live changes. * When a Spot Instance Request is deleted/canceled, an attempt is made to terminate the last-known attached spot instance. Race conditions dictate that this attempt cannot guarantee that the associated spot instance is terminated immediately. Implementation notes: * This version of aws_spot_instance_request borrows a lot of common code from aws_instance. * In order to facilitate borrowing, we introduce `awsInstanceOpts`, an internal representation of instance details that's meant to be shared between resources. The goal here would be to refactor ASG Launch Configurations to use the same struct. * The new aws_spot_instance_request acc. test is passing. * All aws_instance acc. tests remain passing.
2015-06-05 17:12:09 +02:00
Refresh: InstanceStateRefreshFunc(conn, id),
Timeout: 10 * time.Minute,
Delay: 10 * time.Second,
MinTimeout: 3 * time.Second,
}
_, err := stateConf.WaitForState()
if err != nil {
return fmt.Errorf(
"Error waiting for instance (%s) to terminate: %s", id, err)
}
return nil
}
func iamInstanceProfileArnToName(ip *ec2.IamInstanceProfile) string {
if ip == nil || ip.Arn == nil {
return ""
}
parts := strings.Split(*ip.Arn, "/")
return parts[len(parts)-1]
}