Substrate documentation
2023.05
Source & Binary
Search
⌃K
Links

Release notes

2023.05

  • Add support for the AWS_RoleName custom user attribute in Okta, which catches up with Azure AD and Google by allowing Okta users to assign initial AWS IAM roles to each of their users individually.
  • Save a click on subsequent Intranet logins when using a Google IdP and when multiple Google accounts are logged in by remembering the Google Workspace domain in a cookie.
  • Change substrate create-role's canned access policy options from -administrator and -read-only to -administrator-access and -read-only-access to match the AWS-managed policy names and distinguish them from the similarly named -admin account-selection option.
  • When Substrate's run from a Fish shell, environment variables printed by substrate credentials and substrate assume-role will use Fish syntax.
  • Document the install.sh program that's been quietly distributed with Substrate for the past few months. See the documentation on unattended installation for more information.
  • Bug fix: Don't include -admin in the output of substrate roles just because -humans was included in the arguments to substrate create-role. The previous behavior potentially granted too many privileges in the admin account. Re-running substrate create-role with -humans and without -admin will correct the role's privileges in the admin account.
  • Bug fix: Don't attach policies specified by -administrator-access, -read-only-access, -policy-arn, or -policy options to admin accounts when -humans is given unless -admin is given, too. The previous behavior potentially granted too many privileges in the admin account. Re-running substrate create-role will correct the role's privileges in the admin account.
  • Bug fix: Accommodate a new error from S3 in which we cannot set ACLs on newly created buckets.
Before upgrading Substrate, if you use Okta as your identity provider, add the okta.users.read.self scope to your Intranet application and the AWS_RoleName attribute to each of your users by integrating your Okta identity provider again. If you use a different identity provider, no action is necessary.
Upgrade Substrate by running substrate upgrade and following its prompts. If your copy of substrate is writeable, this will be all you need to do to upgrade.
After upgrading Substrate, run substrate create-admin-account -quality <quality> to upgrade the Intranet. This is all that's required but, for good measure, you should run sh <(substrate accounts -format shell -no-apply), review what Terraform plans to do, and then run sh <(substrate accounts -auto-approve -format shell) to apply the changes.

2023.04

  • Cache an access key for the CredentialFactory IAM user (required to make 12-hour sessions) in AWS Secrets Manager to make the Credential Factory, both in the browser and the terminal, substantially faster.
  • Inspect the roles in all your AWS accounts concurrently to make substrate roles substantially faster, too.
  • Print much more helpful errors when the underlying error is simply that there weren't any AWS credentials in the environment.
  • Bug fix: In some situations, substrate roles would write out empty filenames for the -assume-role-policy and -policy options. Now it never does that.
  • Bug fix: substrate roles would never write out any -github-actions options no matter how nicely you asked. It is better behaved now.
  • Bug fix: Really, truly, never ask if you're OK submitting telemetry to Source & Binary outside of the first time you run substrate bootstrap-management-account.
  • Bug fix: The cache of accounts wasn't being cleared early enough such that substrate.accounts.txt could have included stale Substrate version numbers.
Upgrade Substrate by running substrate upgrade and following its prompts. If your copy of substrate is writeable, this will be all you need to do to upgrade.
After upgrading Substrate, run substrate create-admin-account -quality <quality> to upgrade the Intranet. This is all that's required but, for good measure, you should run sh <(substrate accounts -format shell -no-apply), review what Terraform plans to do, and then run sh <(substrate accounts -auto-approve -format shell) to apply the changes.

2023.03

  • Added substrate create-role to allow you to create and manage IAM roles and policies across all your AWS accounts. It supports selecting accounts by domain, environment, and quality and can grant access to selected accounts to humans (via your identity provider), AWS services, GitHub Actions configurations, and arbitrary AWS principals.
  • Added substrate delete-role for thoroughly deleting IAM roles created by substrate create-role across all your AWS accounts.
  • Added substrate roles which inspects all your AWS accounts and can recreate all your previous substrate create-role commands. It's analogous to substrate accounts, including supporting a -format shell option.
  • Tailored the output of substrate whoami to the kind of account so that it doesn't report — alarmingly, at first glance — that certain accounts don't have Domain, Environment, and Quality tags.
  • Only support the -non-interactive, -minimally-interactive, and -fully-interactive flags on bootstrapping and account creation subcommands. They had no purpose in other subcommands but muddled their usage messages.
  • Bug fix: Formally disallow spaces and commas in the names of domains, environments, and qualities.
  • Bug fix: More aggressively garbage-collect expired tokens from half-completed eval $(substrate credentials) invocations in Credential Factory.
Upgrade Substrate by running substrate upgrade and following its prompts. If your copy of substrate is writeable, this will be all you need to do to upgrade.
After upgrading Substrate, you should run sh <(substrate accounts -format shell -no-apply). This Substrate upgrade does not require any Terraform changes to be applied so whether and when to do so is left to you.
This month, Substrate's documentation and release notes are moving to https://docs.src-bin.com/substrate/. Old URLs will redirect to their equivalent on the new subdomain. And, for your troubles, we're pleased to now offer search over the Substrate documentation! Going forward, the new canonical URL for the release notes is http://docs.src-bin.com/substrate/releases.

2023.02

  • Allow adopting an existing account as the Substrate-managed audit account to prevent duplication when bootstrapping Substrate in existing organizations.
  • Make the Substrate-managed CloudTrail configuration optional to prevent very expensive duplication when bootstrapping Substrate in existing organizations.
  • Allow opting out of the Service Control Policy that enforces the uses of the EC2 IMDSv2 to allow more time for legacy workloads to be upgraded.
  • Don't make folks go through the exercise of pairing environments and qualities in substrate bootstrap-network-account in case there's only one valid quality.
  • Stop asking for confirmation before applying Terraform changes in root-modules/network/peering, which are completely generated, very simple, and very safe to make substrate bootstrap-network-account a much more pleasant experience.
  • Rewrite module.substrate.tags to be based on the root-modules filesystem hierarchy instead of an external process so it can be used earlier in the Terraform run.
  • Annotate the Intranet's Accounts page, the output of substrate accounts, and the substrate.accounts.txt file with the version of Substrate last used to manage each account in your organization to make upgrades more observable.
  • Remove long-unused substrate-instance-factory security group from admin accounts. The Instance Factory stopped using this security group over a year ago but if there do happen to still be EC2 instances using this security group, substrate create-admin-account will fail.
  • Enforce CSRF protections in the Intranet more aggressively when the CSRF cookie isn't present, closing the transitional period for this security feature.
  • Introduce the SUBSTRATE_DEBUG_AWS_LOGS and SUBSTRATE_DEBUG_AWS_RETRIES environment variables to support debugging Substrate.
  • Bug fix: Race between VPC-related Terraform resources and sharing/tagging the appropriate VPC in admin and service accounts is no longer a race.
Upgrade Substrate by running substrate upgrade and following its prompts. If your copy of substrate is writeable, this will be all you need to do to upgrade.
After upgrading Substrate, you should run sh <(substrate accounts -format shell -no-apply), review what Terraform plans to do, and then run sh <(substrate accounts -auto-approve -format shell) to apply the changes. Commit the new substrate.enforce-imdsv2 and substrate.manage-cloudtrail files to version control.

2023.01

  • Upgrade Terraform to version 1.3.6 and the Terraform AWS provider to at least version 4.47.0.
  • substrate terraform will now download and unzip the correct version of Terraform into the same directory as substrate itself.
  • substrate intranet-zip now supports -base64sha256 and -format json which are used in the generated Terraform code to prevent Intranet downgrades.
  • Bug fix: substrate -version now writes to standard output instead of standard error to make install automation easier to write.
Upgrade Substrate by running substrate upgrade and following its prompts. If your copy of substrate is writeable, this will be all you need to do to upgrade.
After upgrading Substrate:
  1. 1.
    Upgrade Terraform to version 1.3.6 by running substrate terraform, downloading from Hashicorp, or some other means of your choice.
  2. 2.
    Run sh <(substrate accounts -format shell -no-apply) to review what Terraform plans to do.
  3. 3.
    Run sh <(substrate accounts -auto-approve -format shell) to apply the changes.

2022.12

  • Add support for Azure Active Directory identity providers. See changing identity providers and integrating your Azure AD identity provider if you want to switch.
  • Bug fix: Don't interpret the new default value for -quality as an erroneously user-supplied value with -management or -special in substrate assume-role.
  • Bug fix: Tolerate substrate.qualities being missing when trying to find a suitable default value for -quality options.
  • Bug fix: Change to SUBSTRATE_ROOT, if set, before trying to use substrate.qualities to find a suitable default value for -quality options.
Upgrade Substrate by running substrate upgrade and following its prompts. If your copy of substrate is writeable, this will be all you need to do to upgrade.
After upgrading Substrate, you should run sh <(substrate accounts -format shell -no-apply), review what Terraform plans to do, and then run sh <(substrate accounts -auto-approve -format shell) to apply the changes.

2022.11

  • -quality "..." is now optional in organizations that have only one line in substrate.qualities, which can allow substrate assume-role, substrate create-account, and substrate create-admin-account invocations to be a little shorter. Note that substrate accounts -format shell will continue to generate the longer version with an explicit -quality option.
  • Use only tags to determine whether an account for a given domain, environment, and quality exists. This means that the email address on a Substrate-managed account doesn't have to be derived from the email address on the management account, that you don't absolutely have to change every account's email address if you want to change your management account's email address, and that you don't have to change the email address on every account you invite into your organization in order for Substrate to work with it.
  • Add -number to substrate create-account which allows you to tag an existing organization account with a domain, environment, and quality, generate Terraform modules for it, and manage its basic IAM roles just like any other Substrate-managed account.
  • Expand the bucket policy on the S3 buckets created in the deploy account in each region to allow s3:GetBucketLocation, s3:GetObject*, s3:ListBucket*, and s3:PutObject*.
  • Keep the Intranet's logs in CloudWatch for seven days instead of just one to facilitate debugging after the fact.
  • Transparently switch the Intranet's Lambda functions to run on the ARM architecture, which is 20% cheaper thanks in part to the underlying ARM instances using less power.
  • Bug fix: Correctly detect when an organization hits its limit on the number of accounts it can contain and open a support case to raise that limit.
Upgrade Substrate by running substrate upgrade and following its prompts. If your copy of substrate is writeable, this will be all you need to do to upgrade.
After upgrading Substrate, you should run sh <(substrate accounts -format shell -no-apply), review what Terraform plans to do, and then run sh <(substrate accounts -auto-approve -format shell) to apply the changes.

2022.10

  • Change Substrate's internal use of AWS IAM roles to only assume a role if it's different than the role Substrate has already assumed. This ensures folks won't run afoul of the new, stricter evaluation of IAM roles' trust policies, as outlined in Announcing an update to IAM role trust policy behavior and set to take full effect February 15, 2023.
  • Allow the DeployAdministrator and NetworkAdministrator roles to assume themselves, explicitly allowing this after the new, stricter evaluation of IAM roles' trust policies (described above) takes full effect.
  • Add modules/common/global and modules/common/regional which will be instantiated in every new service account. See below for directions on opting existing service accounts into this new default behavior.
  • Add modules/deploy/global and modules/deploy/regional as blank slates so folks can add resources (e.g. AWS ECR repositories) to their deploy account in every region.
  • Change the schema of the output of substrate credentials -format json to be compatible with the AWS SDK's credential_process directive. Specifically, rename AccessKeyID to AccessKeyId and Expires to Expiration, as documented in Sourcing credentials with an external process.
  • Add substrate credentials -no-open to give folks the opportunity to choose which browser window opens the Credential Factory authorization.
  • Memoize and parallelize listing accounts to improve performance.
  • Memoize OAuth OIDC signing keys in the Intranet to improve Intranet performance.
  • Bug fix: Avoid infinite recursion when using substrate credentials as a credential_process for the AWS CLI/SDK. Substrate never intended to read ~/.aws/config or ~/.aws/credentials but was doing so inadvertantly; it no longer reads either of these files.
If you wish to instantiate the new common modules in your existing service accounts, take the following steps for each domain:
  1. 1.
    Add the following block to modules/domain/global/main.tf:
    module "common" {
    providers = {
    aws = aws
    aws.us-east-1 = aws.us-east-1
    }
    source = "../../common/global"
    }
  2. 2.
    Add the following block to modules/domain/regional/main.tf:
    module "common" {
    providers = {
    aws = aws
    aws.network = aws.network
    }
    source = "../../common/regional"
    }
  3. 3.
    Run substrate create-account -domain`` ``domain``-environment``environment``-quality``quality for each domain service account.
These modules aren't being instantiated in existing service accounts automatically because Substrate can't guarantee that's safe.
Upgrade Substrate by running substrate upgrade and following its prompts. If your copy of substrate is writeable, this will be all you need to do to upgrade.
After upgrading Substrate, you should run sh <(substrate accounts -format shell -no-apply), review what Terraform plans to do, and then run sh <(substrate accounts -auto-approve -format shell) to apply the changes.

2022.09

  • Ensure folks assigned non-Administrator roles in their identity provider can always get credentials and get into the AWS Console. This will manifest as Terraform changing an ARN to "*" in your admin account.
  • Expand the Intranet's logging to CloudWatch Logs to better help folks diagnose any substrate credentials failures they may encounter.
  • Tune retries and timeouts in the Credential Factory and substrate credentials to ensure we get more meaningful error messages and fewer 504 Gateway Timeout responses. This will manifest as Terraform lowering the Intranet Lambda functions' timeouts.
  • Remove the long-unnecessary attachment of arn:aws:iam::aws:policy/service-role/AmazonAPIGatewayPushToCloudWatchLogs to your admin account's Administrator role. This will manifest as Terraform destroying module.intranet.aws_iam_role_policy_attachment.admin-cloudwatch in your admin account.
  • Bug fix: Garbage collect expired tokens that were never fetched by substrate credentials (whether due to bugs, the process being interrupted, or something else).
  • Bug fix: Hide JavaScript paths introduced in 2022.08 from the Intranet's index page.
Get the 2022.09 release by running substrate upgrade and following its prompts. If your copy of substrate is writeable, this will be all you need to do to upgrade.
After upgrading Substrate, you at least need to run substrate create-admin-account -quality`` ``quality to update your Intranet. Even better would be to run sh <(substrate accounts -format shell -no-apply), review what Terraform plans to do, and then run sh <(substrate accounts -auto-approve -format shell) to apply the changes.

2022.08

  • Automate the tedious step of logging out of the AWS Console before logging in again in a different account. The window that opens will now briefly display the AWS homepage before redirecting to the AWS Console. (This marks the first introduction of JavaScript in the Intranet but, having exhausted all other options, this seems a worthwhile cause. Note, however, that this enhancement does not cover the -console option to substrate assume-role in order to keep that command completely decoupled from the Intranet.)
  • Add -no-apply to substrate accounts -format shell to enable you to review all the changes Terraform is planning to make across all your accounts at once. (After you review, you could swap -no-apply for -auto-approve or run terraform apply in each root module directly.)
  • Enforce a one-hour time limit for invocations of substrate credentials. There's a theoretical security benefit to doing so but mostly this prevents forgotten shells from running up your Lambda and CloudWatch bill while muddling your Intranet's logs.
  • Add a new substrate upgrade command, to be put into service with the 2022.09 release. (More details and updated documentation will accompany that release.)
  • Bug fix: Prevent pathologically long email addresses from breaking Credential Factory by exceeding the allowed length of RoleSessionName in the sts:AssumeRole API.
  • Bug fix: Don't fail spuriously when trying to create an account that already exists when the organization is at its limit for accounts.
  • Bug fix: Prevent a rare crash when submitting telemetry to Source & Binary.
Upgrade Substrate as in the updated installation manual:
tar xf substrate-version-commit-OS-ARCH.tar.gz -C ~/bin --strip-components 2 substrate-version-commit-OS-ARCH/bin/substrate
Each released version and commit is offered in four binary formats; choose the appropriate one for your system. OS is one of “darwin” or “linux” and ARCH is one of “amd64” or “arm64”.
You can install Substrate wherever you like. If ~/bin doesn't suit you, just ensure the directory where you install it is on your PATH.
After upgrading Substrate, the best idea is to run sh <(substrate accounts -format shell -no-apply), review what Terraform proposes to do, and then run sh <(substrate accounts -auto-approve -format shell) to ensure your code and your AWS organization don't diverge. If you need a minimal upgrade process, it's substrate create-admin-account -quality`` ``quality to update your Intranet.
Advance notice of an upcoming change: Next month's release will delete an old EC2 security group that was used by the Instance Factory until late 2021. Beware that, if you have any Instance Factory instances from 2021 or earlier, you'll have to change their security group or terminate them before upgrading next month.

2022.07

  • Restructure the Substrate distribution to:
    • Separate the substrate program in bin/ from optional extra programs in opt/bin/.
    • Remove deprecated symbolic links from the distribution (though the functionality has not been and will not be removed from substrate so previously installed symbolic links will continue to function).
    • Include source code in release tarballs to obviate the need to micromanage GitHub collaborators.
  • Always prefix substrate accounts -format shell output with set -e -x so the shell will stop on error and print the commands as they're executed.
  • Add -auto-approve to substrate accounts -format shell to enable one-command, non-interactive upgrades for the most daring among us.
  • Add module.substrate.cidr_prefix to the other outputs of the convenience module.substrate that's automatically available in domain modules.
  • Bug fix: substrate create-account in 2022.06 failed to create accounts but now it's back. (It had one job!)
  • Bug fix: Substrate shell completion never fully worked in Z shell but now it does.
  • Bug fix: Running Lambda function processes could corrupt themselves with environment variables for the CredentialFactory user, losing access to the Intranet role until the process ends.
Upgrade Substrate as in the updated installation manual:
tar xf substrate-version-commit-OS-ARCH.tar.gz -C ~/bin --strip-components 2 substrate-version-commit-OS-ARCH/bin/substrate
Each released version and commit is offered in four binary formats; choose the appropriate one for your system. OS is one of “darwin” or “linux” and ARCH is one of “amd64” or “arm64”.
You can install Substrate wherever you like. If ~/bin doesn't suit you, just ensure the directory where you install it is on your PATH.
After upgrading Substrate:
  1. 1.
    substrate bootstrap-management-account
  2. 2.
    substrate bootstrap-network-account
  3. 3.
    substrate bootstrap-deploy-account
  4. 4.
    substrate create-admin-account -quality`` ``quality for each of your admin accounts
  5. 5.
    substrate create-account -domain`` ``domain``-environment``environment``-quality``quality for each of your service accounts
If your shell supports process substitution, you can run sh <(substrate accounts -format shell) to run all of these, in the proper order, in one command. As of this release you can make this non-interactive as sh <(substrate accounts -auto-approve -format shell) but this is not recommended as it forgoes your opportunity to object before Terraform applies changes.

2022.06

  • Substrate subcommands that create AWS accounts or apply Terraform code (substrate bootstrap-* and substrate create-*account) now prevent Substrate downgrades by checking the tags on the accounts themselves.
  • POST requests to the Instance Factory have long-overdue CSRF mitigations in place.
  • Upgrade Terraform to version 1.2.3 and the Terraform AWS provider to at least version 4.20.0.
  • Address deprecation warnings about aws_subnet_ids everywhere except in the generated modules/peering-connection directory, which will take a little more time.
  • Make the OAuth OIDC flow in your Intranet a little more debuggable in case of IdP misconfiguration or a Substrate bug.
  • Bug fix: Don't give up too early waiting for IAM credentials to become usable. This caused substrate credentials to occasionally hang forever in 2022.05.
After upgrading Substrate:
  1. 1.
    Upgrade to Terraform 1.2.3​
  2. 2.
    substrate bootstrap-management-account
  3. 3.
    substrate bootstrap-network-account
  4. 4.
    substrate bootstrap-deploy-account
  5. 5.
    substrate create-admin-account -quality`` ``quality for each of your admin accounts
  6. 6.
    substrate create-account -domain`` ``domain``-environment``environment``-quality``quality for each of your service accounts
If your shell supports process substitution, you can upgrade Terraform and then run sh <(substrate accounts -format shell) to run all of these, in the proper order, in one command.

2022.05

  • Allow customization of EC2 instances from the Instance Factory by using a launch template named InstanceFactory-arm64 or InstanceFactory-x86_64, if the one matching the requested instance type is defined. See customizing EC2 instances from the Instance Factory for details and an example.
  • Add cloudtrail:DeleteTrail to the (short) list of APIs that are denied by the Substrate-managed service control policy on your management account.
  • Remove verion constraints from Terraform modules in the modules/ tree, instead letting all the version constraints come from the root module.
  • Upgrade the Terraform AWS provider to at least 4.12.
  • Remove dormant copies of various parts of the Intranet, which were deprecated as their functionality was moved into the monolithic Intranet IAM role and Lambda function.
  • Bug fix: Add organizations:DescribeOrganization to the IAM policy attached to the CredentialFactory IAM user so that it can orient itself fully.
  • Bug fix: Submit your Intranet URL to AWS so that when you're logged out of the AWS Console it links you to how you get back in instead of to the Substrate product page.
  • Bug fix: Sort Terraform resources by their type and label as has always been intended. This will result in differences in source code but not in Terraform plans.
  • Bug fix: Print prompts to standard error so they don't corrupt parseable output.
After upgrading Substrate:
  1. 1.
    substrate bootstrap-management-account
  2. 2.
    substrate create-admin-account -quality`` ``quality for each of your admin accounts

2022.04

  • Enforce, via organization-wide Service Control Policy, that EC2 instances must be launched with access to IMDSv2 and not IMDSv1. The Instance Factory has been launching compatible instances since 2021.10. If for some reason you need to roll this step back, use your Intranet's Accounts page to open the AWS Console in your management account with the OrganizationAdministrator role, visit https://console.aws.amazon.com/organizations/v2/home/policies/service-control-policy, and delete SubstrateServiceControlPolicy. When you've migrated whatever needed IMDSv1 to use IMDSv2, re-run substrate bootstrap-management-account.
  • Substrate now ships with a rudimentary autocomplete mechanism for Bash, Z shell, and other shells with compatibility for Bash completion.
  • Substrate can now be used to drive the --profile option to the standard AWS CLI. See using AWS CLI profiles for details.
  • Upgrade the AWS Terraform provider to version 4.9.0 or (slightly) newer.
  • Remove the dependency on the AWS CLI in the generated modules/substrate Terraform code.
  • Bug fix: Lessen the possibility of a TooManyRequestsException from AWS Organizations during Terraform runs.
  • Bug fix: Update the SubstrateVersion tag on your AWS accounts themselves when Substrate tries to create them and finds that they already exist.
  • Bug fix: This time Substrate actually asks if it may post telemetry to Source & Binary as promised in the previous release.
  • Bug fix: Prevent a rare crash when trying to post tememetry early so commands can exit earlier.
  • Bug fix: Use Terraform resource references in root-modules/deploy to avoid a race during substrate bootstrap-deploy-account.
  • Bug fix: Properly support older instance types, especially t2, in the Instance Factory.
After upgrading Substrate:
  1. 1.
    ​Configure Substrate shell completion​
  2. 2.
    substrate bootstrap-management-account
  3. 3.
    substrate create-admin-account -quality`` ``quality for each of your admin accounts

2022.03

  • Substrate now asks if it may post telemetry to Source & Binary. The data will be used to better understand how Substrate is being used and how it can be improved.
  • Address deprecation warnings from the AWS Terraform provider by refactoring root-modules/deploy.
  • Bug fix: Correctly pass an AWS access key to Terraform even if that access key is from an IAM user. This situation is unlikely but can come up during bootstrapping in brownfield environments.
  • Bug fix: Display environments and qualities in the order they're defined in substrate.environments and substrate.qualities on the Intranet's Accounts page and in the output of substrate accounts and substrate root-modules.
After upgrading Substrate:
  1. 1.
    substrate bootstrap-deploy-account
  2. 2.
    substrate-create-admin-account -quality="..." for each of your admin accounts

2022.02

  • Upgrade to Terraform 1.1.6.
  • Manage L-29A0C5DF, the AWS service limit on the number of accounts in an AWS organization, so that substrate-create-admin-account and substrate-create-account can proceed smoothly when you create your 11th account.
  • Remove the SubstrateVersion tag from Terraform-managed resources. It hasn't been as helpful here as it is on Substrate-managed resources. Plus, Terraform plans are much easier to read without it.
  • Bug fix: Pin the Terraform AWS provider to versions less than 4.0 which contains breaking changes that will be addressed in a subsequent Substrate release.
You must upgrade to Terraform 1.1.6 in order to use Substrate 2022.02. Terraform 1.1.6 may be found here:
After upgrading Substrate, do the following to land the Terraform upgrade and remove the SubstrateVersion tags:
  1. 1.
    substrate-bootstrap-network-account
  2. 2.
    substrate-bootstrap-deploy-account
  3. 3.
    substrate-create-admin-account -quality="..." for each of your admin accounts
  4. 4.
    substrate-create-account -domain="..." -environment="..." -quality="..." for each of your service accounts

2022.01

  • The -role="..." option to substrate-assume-role now defaults to OrganizationAdministrator, Auditor, DeployAdministrator, or NetworkAdministrator for the special accounts and Administrator for admin and service accounts (or Auditor pretty much across the board, if you begin in the Auditor role). This should save a great deal of typing.
  • Add a navigational header to the Intranet to help folks get around.
  • Allow the Auditor role in the audit account to use Amazon Athena to query the CloudTrail logs stored in S3 there. See auditing your Substrate-managed AWS organization for more details.
  • Bug fix: Substrate 2021.12 inadvertantly stopped accepting EC2 instance profile credentials on Instance Factory instances. (The instances are still assigned the correct role, Substrate programs just wouldn't use it.) Substrate programs once again use EC2 instance profile credentials when available.
After upgrading Substrate:
  1. 1.
    substrate-create-admin-account -quality="..."
  2. 2.
    Upgrade Substrate in your Instance Factory instances, if you install it there

2021.12

  • Substate now uses your default region (the one you chose to host your organization’s CloudTrail logs, among other things) when it executes global root modules. This allows you to more completely decouple yourself from us-east-1 if you so choose.
  • Bug fix: Allow the Instance Factory to pass any IAM role configured in your IdP on to EC2 instances your non-Administrator users launch.
  • Safety feature: No longer read ~/.aws/credentials under any circumstances. Since we never use this file in a Substrate-managed AWS organization, reading this file can only serve to “cross the streams” with a legacy AWS account.
The upgrade process this month is much more involved that most. As such, we’ll talk in Slack about when you’re going to perform the upgrade to ensure support’s available in the moment.
Before upgrading Substrate, audit your Terraform modules for resources in global modules that aren’t from global AWS serivces by copying the following program to audit.sh in your Substrate repository and running sh audit.sh.
set -e
​
substrate root-modules |
grep "/global\$" |
while read DIRNAME
do
echo "$DIRNAME" >&2
terraform -chdir="$DIRNAME" state pull >"$DIRNAME/audit.tfstate"
grep -F ":us-east-1:" "$DIRNAME/audit.tfstate" || :
rm -f "$DIRNAME/audit.tfstate"
echo >&2
done
Every resource this program identifies needs to be modified before proceeding. The most likely modification is to add provider = aws.us-east-1 to resources in the Terraform code that manages them.
Block all your coworkers from making Terraform changes however you usually do (announcing in Slack, deactivating CI/CD jobs, taking state file locks, etc.) and move your global state files from us-east-1 to your default region by copying the following program to mv-state.sh in your Substrate repository and running sh mv-state.sh.
set -e
​
DEFAULT_REGION="$(cat "substrate.default-region")"
PREFIX="$(cat "substrate.prefix")"
​
if [ "$DEFAULT_REGION" = "us-east-1" ]
then exit # nothing to do
fi
​
eval $(substrate-assume-role -role="DeployAdministrator" -special="deploy")
​
substrate root-modules |
grep "/global\$" |
while read DIRNAME
do
echo "$DIRNAME" >&2
aws s3 cp "s3://$PREFIX-terraform-state-us-east-1/$DIRNAME/terraform.tfstate" "s3://$PREFIX-terraform-state-$DEFAULT_REGION/$DIRNAME/terraform.tfstate"
aws s3 rm "s3://$PREFIX-terraform-state-us-east-1/$DIRNAME/terraform.tfstate"
echo >&2
done
Once you’ve run this program, there’s a provider to thread through the tree of Terraform modules before you can upgrade to Substrate 2021.12.
  • Add the following four lines in the domain module stanzas in root-modules/*/*/*/global/main.tf:
    providers = {
    aws = aws
    aws.us-east-1 = aws.us-east-1
    }
  • Add the following three lines below aws = { in modules/*/global/versions.tf except modules/lambda-function/global/versions.tf, modules/substrate/global/versions.tf, and your own modules:
    configuration_aliases = [
    aws.us-east-1,
    ]
(I regret not being able to provide a patch(1) file for these operations but the contents of versions.tf post-Terraform 1.0 are too unpredictable to do so safely.)
Now you can upgrade Substrate. Don’t release your block just yet, though.
After upgrading Substrate:
  1. 1.
    substrate-bootstrap-deploy-account
  2. 2.
    substrate-create-admin-account -quality="..." for each of your admin accounts
  3. 3.
    substrate-create-account -domain="..." -environment="..." -quality="..." for each of your service accounts
Once all of these have run successfully, ensure all your coworkers upgrade Substrate and unblock Terraform changes.
I regret the complexity of this upgrade process but feel it is, on balance, less risky than attempting to hide all this motion behind automation. Thanks for your patience.

2021.11

  • New installations no longer configure a SAML provider. Instead, all AWS API and Console access is brokered by OAuth OIDC and your Intranet. Existing SAML providers are not removed.
  • For those using Google as their IdP, read the name of the role to assume in the Credential and Instance Factories and the AWS Console from custom attributes on Google Workspace users. (For Okta users, Administrator is still the default.)
  • Forward Cookie headers to HTTP(S) services wired into the Intranet by the experimental new modules/intranet/regional/proxy module. The theoretical security benefit of not exposing raw cookies (and instead exposing identity) is not remotely worth the loss in functionality it cost.
  • Reduce the chance of Intranet misconfiguration by limiting which API Gateways can forward to the substrate-intranet Lambda function.
  • Bug fix: The links to other parts of the Intranet from the page that substrate-credentials opens in your browser are relative links and were missing the ../ prefix, which has now been added.
  • Removed version pinning of the long-unused Terraform archive provider.
  • Removed -no-cloudwatch from substrate-bootstrap-management-account, substrate-create-admin-account, and substrate-create-account in favor of just actually detecting when it’s necessary and not doing it when it’s not.
Before upgrading Substrate, if you’re using Google as your IdP:
  1. 1.
    Add an additional custom attribute as follows:
    1. 1.
      Visit https://admin.google.com/ac/customschema in a browser (or visit https://admin.google.com, click Users, click More, and click Manage custom attributes)
    2. 2.
      Click the AWS section
    3. 3.
      In the blank bottom row, enter “RoleName” for Name, select “Text” for Info type, select “Visible to user and admin” for Visibility, select “Single Value” for No. of values
    4. 4.
      Click SAVE
  2. 2.
    Visit https://admin.google.com/ac/users and set the RoleName attribute in the AWS category to “Administrator” for every user authorized to use AWS.
  3. 3.
    Visit https://console.cloud.google.com/apis/library/admin.googleapis.com, confirm the selected project is the one that contains your Intranet’s OAuth OIDC configuration (its name will be listed next to “Google Cloud Platform” in the header), and click ENABLE.
After upgrading Substrate:
  1. 1.
    Run substrate-create-admin-account -quality="..." to upgrade your Intranet.

2021.10

  • The Intranet’s /accounts page now logs you into the AWS Console and assumes the specified role without requiring you to have already been logged in.
  • The -console option to substrate-assume-role likewise now logs into the AWS Console and assumes the specified role without requiring you to have already been logged in.
  • Auditor roles can now assume other Auditor roles, making it possible for Auditor to move throughout the organization while retaining its read-only status.
  • The lists of principals that can assume the Administrator and Auditor roles may now be augmented by adding a JSON-encoded assume role policy in substrate.Administrator.assume-role-policy.json and/or substrate.Auditor.assume-role-policy.json.
  • Enforce the use of IMDSv2 on instances from the Instance Factory. This is a prerequisite for organization-wide enforcement of using IMDSv2, which is an important default that reduces the potential impact of SSRF vulnerabilities.
  • substrate-whoami output now also includes your IAM role ARN.
  • Prompt folks to cd or set SUBSTRATE_ROOT when they try to eval $(substrate-credentials) from outside the Substrate repository.
  • Allow all accounts in the organization, not just admin accounts, to read shared CloudWatch metrics.
  • Added experimental modules/intranet/regional/proxy that makes it easy to put SSO in front of internal websites and HTTP APIs. See protecting internal websites for more information and an example.
  • Bug fix: Grant s3:PutObjectAcl so that it’s possible for all authorized principals to upload objects with the bucket-owner-full-control canned ACL.
  • Bug fix: Extract substrate-intranet.zip from the substrate binary during Terraform runs in root-modules/admin/*/* instead of only during substrate-create-admin-account. This makes it far less painful for mulitple teammates to work in the same Substrate repository and for CI/CD systems to apply Terraform changes.
  • Bug fix: Prevent a race between VPC sharing and tagging that caused substrate-create-admin-account and substrate-create-account to fail every time they were used to actually create an account.
  • Added -no-cloudwatch to substrate-bootstrap-management-account, substrate-create-admin-account, and substrate-create-account that skips the slow process of managing all the roles necessary for cross-account CloudWatch sharing. (Useful if you’re certain you’ve not created a new account and you’re in a hurry.)
After upgrading Substrate:
  1. 1.
    Run substrate-bootstrap-deploy-account to fix the bucket policy so that all authorized principals in the organization can upload to the deploy artifact bucket(s).
  2. 2.
    Run substrate-create-admin-account -quality="..." to upgrade your Intranet and Auditor roles. Note well this will produce a fair number of new resources; this is step one in a multi-month process of brining some naming consistency to Substrate-managed resources in IAM, Lambda, and other AWS services.

2021.09.28

  • Bug fix: Properly detect when Substrate tools are invoked via symbolic links (i.e. in their original substrate-assume-role form rather than their new substrate assume-role form) on MacOS.
If you’re upgrading from 2021.08, follow the upgrade instructions from 2021.09. If you already upgraded to 2021.09, there are no further upgrade steps.

2021.09

This release changes the interactive interface to substrate-bootstrap-network-account and substrate-create-admin-account to make them easier to run in CI. If you are automating these commands by providing yes and no answers on standard input, this release will break your automation; you should run these commands interactively first to see what’s changed. The details of what’s changed are listed in the usual format below.
  • Move all Substrate commands into the substrate binary with symbolic links replacing the substrate-* binaries from previous releases. This can mostly be considered a no-op but note that now Substrate commands may be also be invoked as substrate`` ``subcommand. This is not a deprecation notice for the original invocation style.
  • Added -fully-interactive, -minimally-interactive, and -non-interactive to all Substrate commands. -fully-interactive is almost identical (see below) to the behavior of 2021.08 and earlier releases. -minimally-interactive is the new default and removes the incessant “is this correct? (yes/no)” dialogs, which I thought would be welcome but turned out to be annoying. -non-interactive will never prompt for input and will instead exit with a non-zero status if input is required.
  • Changed the interactive prompts concerning Google and Okta configuration to make them less bothersome and (in the Okta case) less prone to unintentional changes. If you are automating substrate-create-admin-account by providing yes and no answers on standard input, this change will break your automation; you should run this command interactively first to see what’s changed.
  • Added a confirmation to substrate-create-admin-account and substrate-create-account to prevent errant creation of new AWS accounts (which are tedious to delete in case creation was a mistake) plus a new -create option to suppress that confirmation.
  • Updated the Substrate-managed Service Control Policy attached to your organization to deny access to the cloudtrail:CreateTrail API. Substrate creates a multi-region, organization-wide trail early in its initialization. This policy prevents additional trails from being created because they are excessively expensive and redundant.
  • Added e-mail address columns to tables of AWS accounts in substrate.accounts.txt, substrate-accounts, and the Intranet’s /accounts page.
  • Added -format=shell to substrate-accounts, which enumerates AWS accounts as shell commands to the various substrate-bootstrap-* and substrate-create-* commands. This is useful for driving CI/CD of Terraform changes. It’s also useful for automating Substrate upgrades.
  • Added substrate-root-modules, which enumerates every Substrate-managed Terraform root module in a sensible order. This, too, is useful for driving CI/CD of Terraform changes.
  • Added a new root-modules/deploy/global root module under the management of substrate-bootstrap-deploy-account. Substrate doesn’t manage any resources there but you’re free to.
  • Bug fix: Ensure that objects put into the deploy buckets in S3 by any account in your organization may actually be fetched by other accounts in your organization. Requires objects be uploaded with the bucket-owner-full-control canned ACL.
  • Bug fix: Avoid a fork bomb in case substrate-whoami is invoked with a / in the pathname (i.e. as ~/bin/substrate-whoami).
After upgrading Substrate:
  1. 1.
    Run substrate-bootstrap-management-account to update your organization’s Service Control Policy.
  2. 2.
    Run substrate-bootstrap-deploy-account to reconfigure the deploy buckets in S3 and generate the global root module.
  3. 3.
    Run substrate-create-admin-account -quality="..." to add the e-mail address column to your Intranet’s /accounts page.

2021.08

  • Roll substrate-apigateway-authorizer, substrate-credential-factory, and substrate-instance-factory into substrate-intranet. This is a no-op listed here for transparency. It’s a prerequisite step towards unifying all the Substrate tools as subcommands of substrate, thereby reducing the size and complexity of the Substrate distribution.
  • Stop using the archive and external Terraform providers by embedding substrate-intranet.zip directly in substrate-create-admin-account. Dependence on these providers will be made optional in a subsequent release.
  • VPCs are no longer shared organization-wide, leaving the fine-grained VPC sharing introduced in 2021.07 to maintain each service account’s access to its intended VPCs.
  • The Instance Factory now supports ARM instances (i.e. the a1, c6g, m6g, r6g, and t4g families).
  • Bug fix: Switch back to the original working directory in substrate-assume-role (which will have changed if invoked with SUBSTRATE_ROOT set) before forking and executing a child process.
  • Added substrate-whoami to make it easy to learn the domain, environment, and quality of the AWS account your current credentials operate on.
  • Added -format=json to substrate-accounts to make it easier to enumerate and act programatically on every AWS account in your organization. See enumerating all your AWS accounts for an example.
After upgrading Substrate:
  1. 1.
    Run substrate-bootstrap-management-account to grant substrate-whoami the permissions it needs.
  2. 2.
    Run substrate-bootstrap-network-account to remove coarse-grained organization-wide VPC sharing.
  3. 3.
    Run substrate-create-admin-account -quality="..." to upgrade your Intranet.

2021.07

  • The Intranet’s /accounts page now opens the AWS Console in new browser tabs as it probably always should have.
  • Substrate now only manages the version constraint on the archive, aws, and external providers rather than all of versions.tf. This opens the door to Substrate users adding (and version constraining) additional providers. See additional Terraform providers for an example.
  • Upgrade to and pin Terraform 1.0.2 and the aws provider >= 3.49.0.
  • Tag many more AWS resources with Manager and SubstrateVersion using the default_tags facility of the AWS provider. If you encounter the following error, remove Manager and SubstrateVersion (if present) from the indicated resources and re-run.\
    Error: "tags" are identical to those in the "default_tags" configuration block of the provider: please de-duplicate and try again
  • All Substrate tools will now change their working directory to the value of the SUBSTRATE_ROOT environment variable, if set, rather than always proceeding in whatever the working directory was when invoked.
  • Share VPCs specifically with accounts that match their environment and quality. This is a no-op that enables a future release to remove organization-wide VPC sharing.
You must upgrade to Terraform 1.0.2 in order to use Substrate 2021.07. Terraform 1.0.2 may be found here:
After upgrading Terraform and Substrate:
  1. 1.
    Run substrate-bootstrap-network-account and substrate-bootstrap-deploy-account to complete the Terraform 1.0.2 upgrade there. Note well that tags and tags_all output will be somewhat confusing but will ultimately do the right thing.
  2. 2.
    Run substrate-create-admin-account and substrate-create-account to complete the Terraform 1.0.2 upgrade for each of your admin and service accounts. Here, too, note well that tags and tags_all output will be somewhat confusing but will ultimately do the right thing.

2021.06

  • List all the Intranet resources on the Intranet homepage, not just top-level resources.
  • Roll substrate-apigateway-authenticator and substrate-apigateway-index into substrate-intranet. This is a no-op listed here for transparency.
  • Tag shared VPCs in service accounts to clearly indicate in the AWS Console which one you should be using. This lays the groundwork for finer-grained VPC sharing in a future release.
  • Upgrade to and pin Terraform 0.15.5 and newer providers with relaxed >= version constraints on providers (but not Terraform itself).
You must upgrade to Terraform 0.15.5 in order to use Substrate 2021.06. Terraform 0.15.5 may be found here:
After upgrading Terraform and Substrate:
  1. 1.
    Run substrate-bootstrap-network-account and substrate-bootstrap-deploy-account to complete the Terraform 0.15.5 upgrade there.
  2. 2.
    Run substrate-create-admin-account -quality="..." to update your Intranet.
  3. 3.
    Run substrate-create-account -domain="..." -environment="..." -quality="..." for all your service accounts to tag your shared VPCs.
If you’ve added any stub provider blocks to your modules, leave them in place for now and accept the deprecation warning. Terraform only allows one required_providers block and that is now managed by Substrate. A future release will accommodate these additional providers.

2021.05

  • Bug fix: S3 traffic from private subnets is now correctly routed via the VPC Endpoint and not through the NAT Gateway.
  • Bug fix: Allow outbound IPv6 traffic from Instance Factory instances to match IPv4 and enable use of the IPv6 Internet.
  • Bug fix: Instance Factory instances now have 100GB disks instead of whatever the AMI happened to request, which recently shrank from 8GB to 2GB.
  • Bug fix: The Instance Factory now only lists instances which belong to you. It previously listed all instances for all your users.
  • By popular request, authorize admin accounts to directly access CloudWatch logs and metrics from all your accounts.
After upgrading:
  • Run substrate-bootstrap-network-account to fix S3 routes.
  • Run substrate-create-admin-account -quality=<quality> to enable direct CloudWatch access and make Instance Factory improvements.

2021.04

  • Added /accounts to the Intranet with links to assume the Administrator and Auditor roles in all your accounts in the AWS Console.
  • Added -console to substrate-assume-role which attempts to open the AWS Console’s role switching screen in your web browser with all the values filled in.
  • Added substrate-create-terraform-module which creates the directory structure (with the global and regional pattern), providers, and Substrate metadata for a new Terraform module.
  • Now building for M1 Macs, too.
After upgrading, run substrate-create-admin-account -quality=<quality> to add /accounts to your Intranet.

2021.03

  • Extended AWS Console sessions to 12 hours for organizations using Google as their IdP.
  • Upgrade to and pin Terraform 0.14.7.
  • substrate-bootstrap-network-account now creates peering connections between all VPCs in all regions for each environment across all valid qualities.
  • Fixed a bug in the Administrator role in admin accounts that prevented Instance Factory instances from seamlessly assuming the role.
You must upgrade to Terraform 0.14.7 in order to use Substrate 2021.03. Terraform 0.14.7 may be found here:
After upgrading:
  1. 1.
    rm -f -r root-modules/network/*/peering and remove these files from version control.
  2. 2.
    substrate-bootstrap-network-account to peer all your VPCs that should be peered.
  3. 3.
    substrate-create-admin-account -quality="..." to fix Instance Factory IAM roles, following the Google SAML setup guide if Google is your IdP to also get 12-hour AWS Console sessions.

2021.02

  • Added -format from substrate-credentials to substrate-assume-role per request from a customer. Now credentials can be had with or without the export prefix or as JSON a la aws sts assume-role itself.
  • Removed root-modules/admin/*’s awkward dependency on finding GOBIN in the environment. The generated Makefile in each root module remains, however.
  • Upgrade to and pin Terraform 0.13.6 and the Terraform AWS provider 3.26.0.
  • substrate-assume-role and substrate-credentials now (better) tolerate being invoked from subdirectories of your Substrate repository.
  • Fix a bug in Terraform module generation in which the aws.network provider was incorrectly added to global modules and thus should have been expected in module stanzas.
  • Stop printing the export AWS_ACCESS_KEY_ID=... line when substrate-assume-role is given a command to execute directly.
  • Provide the Intranet’s REST API ID, root resource ID, and a few other necessities as outputs from the intranet/regional module to facilitate adding more resources to these APIs.
You must upgrade to Terraform 0.13.6 in order to use Substrate 2021.02. Terraform 0.13.6 may be found here:

2021.01

You must run substrate-create-admin-account for each of your admin accounts before you'll be able to use eval $(substrate-credentials) to streamline your use of the Credential Factory.

2020.12

You must run substrate-bootstrap-management-account in order to re-tag your former master account as your management account. (This rename follows AWS' own renaming.)

2020.11 and prior releases

Contact [email protected] for prior release notes.
Previous
Welcome
Next - Bootstrapping your Substrate-managed AWS organization
Overview
Last modified 25d ago