Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Teleport 12 Test Plan #20132

Closed
russjones opened this issue Jan 12, 2023 · 34 comments
Closed

Teleport 12 Test Plan #20132

russjones opened this issue Jan 12, 2023 · 34 comments
Labels
test-plan A list of tasks required to ship a successful product release.

Comments

@russjones
Copy link
Contributor

russjones commented Jan 12, 2023

Manual Testing Plan

Below are the items that should be manually tested with each release of Teleport.
These tests should be run on both a fresh installation of the version to be released
as well as an upgrade of the previous version of Teleport.

  • Adding nodes to a cluster @codingllama

    • Adding Nodes via Valid Static Token
    • Adding Nodes via Valid Short-lived Tokens
    • Adding Nodes via Invalid Token Fails
    • Revoking Node Invitation
  • Labels @nklaassen

    • Static Labels
    • Dynamic Labels
  • Trusted Clusters @espadolini

    • Adding Trusted Cluster Valid Static Token
    • Adding Trusted Cluster Valid Short-lived Token
    • Adding Trusted Cluster Invalid Token
    • Removing Trusted Cluster
    • Changing role map of existing Trusted Cluster
  • RBAC @fspmarshall

    Make sure that invalid and valid attempts are reflected in audit log.

    • Successfully connect to node with correct role
    • Unsuccessfully connect to a node in a role restricting access by label
    • Unsuccessfully connect to a node in a role restricting access by invalid SSH login
    • Allow/deny role option: SSH agent forwarding
    • Allow/deny role option: Port forwarding
    • Allow/deny role option: SSH file copying
  • Verify that custom PAM environment variables are available as expected. @rosstimothy

  • Users @tobiaszheller

    With every user combination, try to login and signup with invalid second
    factor, invalid password to see how the system reacts.

    WebAuthn in the release tsh binary is implemented using libfido2 for
    linux/macOS. Ask for a statically built pre-release binary for realistic
    tests. (tsh fido2 diag should work in our binary.) Webauthn in Windows
    build is implemented using webauthn.dll. (tsh webauthn diag with
    security key selected in dialog should work.)

    Touch ID requires a signed tsh, ask for a signed pre-release binary so you
    may run the tests.

    Windows Webauthn requires Windows 10 19H1 and device capable of Windows
    Hello.

    • Adding Users Password Only

    • Adding Users OTP

    • Adding Users WebAuthn

      • macOS/Linux
      • Windows
    • Adding Users via platform authenticator

      • Touch ID
      • Windows Hello
    • Managing MFA devices

      • Add an OTP device with tsh mfa add
      • Add a WebAuthn device with tsh mfa add
        • macOS/Linux
        • Windows
      • Add platform authenticator device with tsh mfa add
        • Touch ID
        • Windows Hello
      • List MFA devices with tsh mfa ls
      • Remove an OTP device with tsh mfa rm
      • Remove a WebAuthn device with tsh mfa rm
      • Attempt removing the last MFA device on the user
        • with second_factor: on in auth_service, should fail
        • with second_factor: optional in auth_service, should succeed
    • Login Password Only

    • Login with MFA

      • Add an OTP, a WebAuthn and a Touch ID/Windows Hello device with tsh mfa add
      • Login via OTP
      • Login via WebAuthn
        • macOS/Linux
        • Windows
      • Login via platform authenticator
        • Touch ID
        • Windows Hello
      • Login via WebAuthn using an U2F device

      U2F devices must be registered in a previous version of Teleport.

      Using Teleport v9, set auth_service.authentication.second_factor = u2f,
      restart the server and then register an U2F device (tsh mfa add). Upgrade
      the installation to the current Teleport version (one major at a time) and try to
      log in using the U2F device as your second factor - it should work.

    • Login OIDC

    • Login SAML

    • Login GitHub

    • Deleting Users

  • Backends @timothyb89

    • Teleport runs with etcd
    • Teleport runs with dynamodb
    • Teleport runs with SQLite
    • Teleport runs with Firestore
  • Session Recording @strideynet

    • Session recording can be disabled
    • Sessions can be recorded at the node
      • Sessions in remote clusters are recorded in remote clusters
    • Sessions can be recorded at the proxy
      • Sessions on remote clusters are recorded in the local cluster
      • With an OpenSSH server without a Teleport CA signed host certificate:
        • Host key checking enabled rejects connection
        • Host key checking disabled allows connection
  • Enhanced Session Recording @jakule

    • disk, command and network events are being logged.
    • Recorded events can be enforced by the enhanced_recording role option.
    • Enhanced session recording can be enabled on CentOS 7 with kernel 5.8+.
  • Restricted Session @jakule

    • Network request are allowed when a policy allow them.
    • Network request are blocked when a policy deny them.
  • Audit Log @Joerger

    • Failed login attempts are recorded

    • Interactive sessions have the correct Server ID

      • server_id is the ID of the node in "session_recording: node" mode
      • server_id is the ID of the node in "session_recording: proxy" mode
      • forwarded_by is the ID of the proxy in "session_recording: proxy" mode

      Node/Proxy ID may be found at /var/lib/teleport/host_uuid in the
      corresponding machine.

      Node IDs may also be queried via tctl nodes ls.

    • Exec commands are recorded

    • scp commands are recorded

    • Subsystem results are recorded

      Subsystem testing may be achieved using both
      Recording Proxy mode
      and
      OpenSSH integration.

      Assuming the proxy is proxy.example.com:3023 and node1 is a node running
      OpenSSH/sshd, you may use the following command to trigger a subsystem audit
      log:

      sftp -o "ProxyCommand ssh -o 'ForwardAgent yes' -p 3023 %[email protected] -s proxy:%h:%p" root@node1
  • Interact with a cluster using tsh @capnspacehook

    These commands should ideally be tested for recording and non-recording modes as they are implemented in a different ways.

    • tsh ssh <regular-node>
    • tsh ssh <node-remote-cluster>
    • tsh ssh -A <regular-node>
    • tsh ssh -A <node-remote-cluster>
    • tsh ssh <regular-node> ls
    • tsh ssh <node-remote-cluster> ls
    • tsh join <regular-node>
    • tsh join <node-remote-cluster>
    • tsh play <regular-node>
    • tsh play <node-remote-cluster>
    • tsh scp <regular-node>
    • tsh scp <node-remote-cluster>
    • tsh ssh -L <regular-node>
    • tsh ssh -L <node-remote-cluster>
    • tsh ls
    • tsh clusters
  • Interact with a cluster using ssh @capnspacehook
    Make sure to test both recording and regular proxy modes.

    • ssh <regular-node>
    • ssh <node-remote-cluster>
    • ssh -A <regular-node>
    • ssh -A <node-remote-cluster>
    • ssh <regular-node> ls
    • ssh <node-remote-cluster> ls
    • scp <regular-node>
    • scp <node-remote-cluster>
    • ssh -L <regular-node>
    • ssh -L <node-remote-cluster>
  • Verify proxy jump functionality @atburke
    Log into leaf cluster via root, shut down the root proxy and verify proxy jump works.

    • tls routing disabled
      • tsh ssh -J <leaf.proxy.example.com:3023>
      • ssh -J <leaf.proxy.example.com:3023>
    • tls routing enabled
      • tsh ssh -J <leaf.proxy.example.com:3080>
      • tsh proxy ssh -J <leaf.proxy.example.com:3080>
  • tsh CA loading @lxea

    Create a trusted cluster pair with a node in the leaf cluster. Log into the root cluster.

    • load_all_cas on the root auth server is false (default) -
      tsh ssh leaf.node.example.com results in access denied.
    • load_all_cas on the root auth server is true - tsh ssh leaf.node.example.com
      succeeds.
  • X11 Forwarding @Joerger

    • Install xeyes and xclip:
      • Linux: apt install x11-apps xclip
      • Mac: Install and launch XQuartz which comes with xeyes. Then brew install xclip.
    • Enable X11 forwarding for a Node running as root: ssh_service.x11.enabled = yes
    • Successfully X11 forward as both root and non-root user
      • tsh ssh -X user@node xeyes
      • tsh ssh -X root@node xeyes
    • Test untrusted vs trusted forwarding
      • tsh ssh -Y server01 "echo Hello World | xclip -sel c && xclip -sel c -o" should print "Hello World"
      • tsh ssh -X server01 "echo Hello World | xclip -sel c && xclip -sel c -o" should fail with "BadAccess" X error

User accounting @tigrato

  • Verify that active interactive sessions are tracked in /var/run/utmp on Linux.
  • Verify that interactive sessions are logged in /var/log/wtmp on Linux.

Combinations @capnspacehook

For some manual testing, many combinations need to be tested. For example, for
interactive sessions the 12 combinations are below.

  • Connect to a OpenSSH node in a local cluster using OpenSSH.
  • Connect to a OpenSSH node in a local cluster using Teleport.
  • Connect to a OpenSSH node in a local cluster using the Web UI. #20831
  • Connect to a Teleport node in a local cluster using OpenSSH.
  • Connect to a Teleport node in a local cluster using Teleport.
  • Connect to a Teleport node in a local cluster using the Web UI.
  • Connect to a OpenSSH node in a remote cluster using OpenSSH.
  • Connect to a OpenSSH node in a remote cluster using Teleport.
  • Connect to a OpenSSH node in a remote cluster using the Web UI.
  • Connect to a Teleport node in a remote cluster using OpenSSH.
  • Connect to a Teleport node in a remote cluster using Teleport.
  • Connect to a Teleport node in a remote cluster using the Web UI.

Teleport with EKS/GKE @AntonAM

  • Deploy Teleport on a single EKS cluster
  • Deploy Teleport on two EKS clusters and connect them via trusted cluster feature
  • Deploy Teleport Proxy outside GKE cluster fronting connections to it (use this script to generate a kubeconfig)
  • Deploy Teleport Proxy outside EKS cluster fronting connections to it (use this script to generate a kubeconfig)

Teleport with multiple Kubernetes clusters @AntonAM

Note: you can use GKE or EKS or minikube to run Kubernetes clusters.
Minikube is the only caveat - it's not reachable publicly so don't run a proxy there.

  • Deploy combo auth/proxy/kubernetes_service outside a Kubernetes cluster, using a kubeconfig
    • Login with tsh login, check that tsh kube ls has your cluster
    • Run kubectl get nodes, kubectl exec -it $SOME_POD -- sh
    • Verify that the audit log recorded the above request and session
  • Deploy combo auth/proxy/kubernetes_service inside a Kubernetes cluster
    • Login with tsh login, check that tsh kube ls has your cluster
    • Run kubectl get nodes, kubectl exec -it $SOME_POD -- sh
    • Verify that the audit log recorded the above request and session
  • Deploy combo auth/proxy_service outside the Kubernetes cluster and kubernetes_service inside of a Kubernetes cluster, connected over a reverse tunnel
    • Login with tsh login, check that tsh kube ls has your cluster
    • Run kubectl get nodes, kubectl exec -it $SOME_POD -- sh
    • Verify that the audit log recorded the above request and session
  • Deploy a second kubernetes_service inside another Kubernetes cluster, connected over a reverse tunnel
    • Login with tsh login, check that tsh kube ls has both clusters
    • Switch to a second cluster using tsh kube login
    • Run kubectl get nodes, kubectl exec -it $SOME_POD -- sh on the new cluster
    • Verify that the audit log recorded the above request and session
  • Deploy combo auth/proxy/kubernetes_service outside a Kubernetes cluster, using a kubeconfig with multiple clusters in it
    • Login with tsh login, check that tsh kube ls has all clusters
  • Test Kubernetes screen in the web UI (tab is located on left side nav on dashboard):
    • Verify that all kubes registered are shown with correct name and labels
    • Verify that clicking on a rows connect button renders a dialogue on manual instructions with Step 2 login value matching the rows name column
    • Verify searching for name or labels in the search bar works
    • Verify you can sort by name colum
  • Test Kubernetes exec via WebSockets - client

Kubernetes auto-discovery @tigrato

  • Test Kubernetes auto-discovery:
    • Verify that Azure AKS clusters are discovered and enrolled for different Azure Auth configs:
      • Local Accounts only
      • Azure AD
      • Azure RBAC
    • Verify that AWS EKS clusters are discovered and enrolled
    • Verify that GCP GKE clusters are discovered and enrolled
  • Verify dynamic registration.
    • Can register a new Kubernetes cluster using tctl create.
    • Can update registered Kubernetes cluster using tctl create -f.
    • Can delete registered Kubernetes cluster using tctl rm.

Kubernetes Secret Storage @tigrato

  • Kubernetes Secret storage for Agent's Identity
    • Install Teleport agent with a short-lived token
      • Validate if the Teleport is installed as a Kubernetes Statefulset
      • Restart the agent after token TTL expires to see if it reuses the same identity.
    • Force cluster CA rotation

Kubernetes Pod RBAC @tigrato

  • Check if Roles prior to v6 are allowed to access all pods.
  • Verify the following scenarios for kubernetes_resources:
    • {"kind":"pod","name":"*","namespace":"*"} - must allow access to every pod.
    • {"kind":"pod","name":"<somename>","namespace":"*"} - must allow access to pod <somename> in every namespace.
    • {"kind":"pod","name":"*","namespace":"<somenamespace>"} - must allow access to any pod in <somenamespace> namespace.
    • Verify support for * wildcards - <some-name>-* and regex for name and namespace fields.
    • Verify support for delete pods collection - must use go-client.
  • Verify scenarios with multiple roles defining kubernetes_resources:
    • Validate that the returned list of pods is the union of every role.
    • Validate that access to other pods is denied by RBAC.
    • Validate that the Kubernetes Groups/Users are correctly selected depending on the role that applies to the pod.
      • Test with a kubernetes_groups that denies exec into a pod
  • Verify the following scenarios for Resource Access Requests to Pods:
    • Create a valid resource access request and validate if access to other pods is denied.
    • Validate if creating a resource access request with Kubernetes resources denied by search_as_roles is not allowed.

Teleport with FIPS mode @r0mant

  • Perform trusted clusters, Web and SSH sanity check with all teleport components deployed in FIPS mode.

ACME @mdwn

  • Teleport can fetch TLS certificate automatically using ACME protocol.

Migrations @r0mant @zmb3

  • Migrate trusted clusters from 2.4.0 to 2.5.0
    • Migrate auth server on main cluster, then rest of the servers on main cluster
      SSH should work for both main and old clusters
    • Migrate auth server on remote cluster, then rest of the remote cluster
      SSH should work

Command Templates

When interacting with a cluster, the following command templates are useful:

OpenSSH

# when connecting to the recording proxy, `-o 'ForwardAgent yes'` is required.
ssh -o "ProxyCommand ssh -o 'ForwardAgent yes' -p 3023 %[email protected] -s proxy:%h:%p" \
  node.example.com

# the above command only forwards the agent to the proxy, to forward the agent
# to the target node, `-o 'ForwardAgent yes'` needs to be passed twice.
ssh -o "ForwardAgent yes" \
  -o "ProxyCommand ssh -o 'ForwardAgent yes' -p 3023 %[email protected] -s proxy:%h:%p" \
  node.example.com

# when connecting to a remote cluster using OpenSSH, the subsystem request is
# updated with the name of the remote cluster.
ssh -o "ProxyCommand ssh -o 'ForwardAgent yes' -p 3023 %[email protected] -s proxy:%h:%[email protected]" \
  node.foo.com

Teleport

# when connecting to a OpenSSH node, remember `-p 22` needs to be passed.
tsh --proxy=proxy.example.com --user=<username> --insecure ssh -p 22 node.example.com

# an agent can be forwarded to the target node with `-A`
tsh --proxy=proxy.example.com --user=<username> --insecure ssh -A -p 22 node.example.com

# the --cluster flag is used to connect to a node in a remote cluster.
tsh --proxy=proxy.example.com --user=<username> --insecure ssh --cluster=foo.com -p 22 node.foo.com

Teleport with SSO Providers @camscale

  • Azure Active Directory (AD) install instructions work
    • Azure Active Directory (AD) Screenshots are up-to-date
  • ActiveDirectory (ADFS) install instructions work
    • Active Directory (ADFS) Screenshots are up-to-date
  • Google Workspace install instructions work @AntonAM
    • Google Workspace Screenshots are up-to-date
  • GitLab install instructions work @capnspacehook
    • GitLab Screenshots are up-to-date
  • OneLogin install instructions work
    • OneLogin Screenshots are up-to-date
  • OIDC install instructions work
    • OIDC Screenshots are up-to-date
  • Okta install instructions work
    • Okta Screenshots are up-to-date
  • All providers with guides in docs are covered in this test plan

GitHub External SSO @Tener

  • Teleport OSS
    • GitHub organization without external SSO succeeds
    • GitHub organization with external SSO fails
  • Teleport Enterprise
    • GitHub organization without external SSO succeeds
    • GitHub organization with external SSO succeeds

tctl sso family of commands @Tener

For help with setting up sso connectors, check out the Quick GitHub/SAML/OIDC Setup Tips

tctl sso configure helps to construct a valid connector definition:

  • tctl sso configure github ... creates valid connector definitions
  • tctl sso configure oidc ... creates valid connector definitions
  • tctl sso configure saml ... creates valid connector definitions

tctl sso test test a provided connector definition, which can be loaded from
file or piped in with tctl sso configure or tctl get --with-secrets. Valid
connectors are accepted, invalid are rejected with sensible error messages.

  • Connectors can be tested with tctl sso test.
    • GitHub
    • SAML
    • OIDC
      • Google Workspace
      • Non-Google IdP

Teleport Plugins @greedy52

  • Test receiving a message via Teleport Slackbot
  • Test receiving a new Jira Ticket via Teleport Jira

AWS Node Joining @gabrielcorado

Docs

  • On EC2 instance with ec2:DescribeInstances permissions for local account:
    TELEPORT_TEST_EC2=1 go test ./integration -run TestEC2NodeJoin
  • On EC2 instance with any attached role:
    TELEPORT_TEST_EC2=1 go test ./integration -run TestIAMNodeJoin
  • EC2 Join method in IoT mode with node and auth in different AWS accounts
  • IAM Join method in IoT mode with node and auth in different AWS accounts

Kubernetes Node Joining @gabrielcorado

  • Join a Teleport node running in the same Kubernetes cluster via a Kubernetes ProvisionToken

Cloud Labels @GavinFrazar

  • Create an EC2 instance with tags in instance metadata enabled
    and with tag foo: bar. Verify that a node running on the instance has label
    aws/foo=bar.
  • Create an Azure VM with tag foo: bar. Verify that a node running on the
    instance has label azure/foo=bar.

Passwordless @codingllama

This feature has additional build requirements, so it should be tested with a pre-release build from Drone (eg: https://get.gravitational.com/teleport-v10.0.0-alpha.2-linux-amd64-bin.tar.gz).

This sections complements "Users -> Managing MFA devices". tsh binaries for
each operating system (Linux, macOS and Windows) must be tested separately for
FIDO2 items.

  • Diagnostics

    Commands should pass all tests.

    • tsh fido2 diag (macOS/Linux)
    • tsh touchid diag (macOS only)
    • tsh webauthnwin diag (Windows only)
  • Registration

    • Register a passworldess FIDO2 key (tsh mfa add, choose WEBAUTHN and
      passwordless)
      • macOS/Linux
      • Windows
    • Register a platform authenticator
      • Touch ID credential (tsh mfa add, choose TOUCHID)
      • Windows hello credential (tsh mfa add, choose WEBAUTHN and
        passwordless)
  • Login

    • Passwordless login using FIDO2 (tsh login --auth=passwordless)
      • macOS/Linux
      • Windows
    • Passwordless login using platform authenticator (tsh login --auth=passwordless)
      • Touch ID
      • Windows Hello
    • tsh login --auth=passwordless --mfa-mode=cross-platform uses FIDO2
      • macOS/Linux
      • Windows
    • tsh login --auth=passwordless --mfa-mode=platform uses platform authenticator
      • Touch ID
      • Windows Hello
    • tsh login --auth=passwordless --mfa-mode=auto prefers platform authenticator
      • Touch ID
      • Windows Hello
    • Passwordless disable switch works
      (auth_service.authentication.passwordless = false)
    • Cluster in passwordless mode defaults to passwordless
      (auth_service.authentication.connector_name = passwordless)
    • Cluster in passwordless mode allows MFA login
      (tsh login --auth=local)
  • Touch ID support commands

    • tsh touchid ls works
    • tsh touchid rm works (careful, may lock you out!)

Device Trust @sfreiberg

Device Trust requires Teleport Enterprise.

This feature has additional build requirements, so it should be tested with a
pre-release build from Drone (eg:
https://get.gravitational.com/teleport-v10.0.0-alpha.2-linux-amd64-bin.tar.gz).

Client-side enrollment requires a signed tsh for macOS, make sure to use the
tsh binary from tsh.app.

A simple formula for testing device authorization is:

# Before enrollment.
# Replace with other kinds of access, as appropriate (db, kube, etc)
tsh ssh node-that-requires-device-trust
> ERROR: ssh: rejected: administratively prohibited (unauthorized device)
# Register the device.
# Get the serial number from "Apple -> About This Mac".
tctl devices add --os=macos --asset-tag=<SERIAL_NUMBER> --enroll
# Enroll the device.
tsh device enroll --token=<TOKEN_FROM_COMMAND_ABOVE>
tsh logout; tsh login
# After enrollment
tsh ssh node-that-requires-device-trust
> $
  • Inventory management

    • Add device (tctl devices add)
    • Add device and create enrollment token (tctl devices add --enroll)
    • List devices (tctl devices ls)
    • Remove device using device ID (tctl devices rm)
    • Remove device using asset tag (tctl devices rm)
    • Create enrollment token using device ID (tctl devices enroll)
    • Create enrollment token using asset tag (tctl devices enroll)
  • Device enrollment

    • Enroll device on macOS (tsh device enroll)

    • Verify device extensions on TLS certificate

      Note that different accesses have different certificates (Database, Kube,
      etc).

      $ openssl x509 -noout -in ~/.tsh/keys/zarquon/llama-x509.pem -nameopt sep_multiline -subject | grep 1.3.9999.3
      > 1.3.9999.3.1=6e60b9fd-1e3e-473d-b148-27b4f158c2a7
      > 1.3.9999.3.2=AAAAAAAAAAAA
      > 1.3.9999.3.3=661c9340-81b0-4a1a-a671-7b1304d28600
    • Verify device extensions on SSH certificate

      ssh-keygen -L -f ~/.tsh/keys/zarquon/llama-ssh/zarquon-cert.pub | grep teleport-device-
      teleport-device-asset-tag ...
      teleport-device-credential-id ...
      teleport-device-id ...
  • Device authorization

    • device_trust.mode other than "off" or "" not allowed (OSS)

    • device_trust.mode="off" doesn't impede access (Enterprise and OSS)

    • device_trust.mode="optional" doesn't impede access, but issues device
      extensions on login

    • device_trust.mode="required" enforces enrolled devices

    • device_trust.mode="required" is enforced by processes, and not only by
      Auth APIs

      Testing this requires issuing a certificate without device extensions
      (mode="off"), then changing the cluster configuration to mode="required" and
      attempting to access a process directly, without a login attempt.

    • Device authorization works correctly for both require_session_mfa=false
      and require_session_mfa=true

    • Device authorization applies to SSH access (all items above)

    • Device authorization applies to Trusted Clusters (root with
      mode="optional" and leaf with mode="required")

    • Device authorization applies to Database access (all items above) @smallinsky

    • Device authorization applies to Kubernetes access (all items above) @tigrato

  • Device audit (see lib/events/codes.go)

    • Inventory management actions issue events (success only)
    • Device enrollment issues device event (any outcomes)
    • Device authorization issues device event (any outcomes)
    • Events with UserMetadata contain TrustedDevice
      data (for certificates with device extensions)

Hardware Key Support @Joerger

Hardware Key Support is an Enterprise feature and is not available for OSS.

You will need a YubiKey 4.3+ to test this feature.

This feature has additional build requirements, so it should be tested with a pre-release build from Drone (eg: https://get.gravitational.com/teleport-ent-v11.0.0-alpha.2-linux-amd64-bin.tar.gz).

Server Access @Joerger

These tests should be carried out sequentially. tsh tests should be carried out on Linux, MacOS, and Windows.

  1. tsh login as user with Webauthn login and no hardware key requirement.
  2. Request a role with role.role_options.require_session_mfa: hardware_key - tsh login --request-roles=hardware_key_required
  • Assuming the role should force automatic re-login with yubikey
  • tsh ssh
    • Requires yubikey to be connected for re-login
    • Prompts for per-session MFA
  1. Request a role with role.role_options.require_session_mfa: hardware_key_touch - tsh login --request-roles=hardware_key_touch_required
  • Assuming the role should force automatic re-login with yubikey
    • Prompts for touch if not cached (last touch within 15 seconds)
  • tsh ssh
    • Requires yubikey to be connected for re-login
    • Prompts for touch if not cached
  1. tsh logout and tsh login as the user with no hardware key requirement.
  2. Upgrade auth settings to auth_service.authentication.require_session_mfa: hardware_key
  • Using the existing login session (tsh ls) should force automatic re-login with yubikey
  • tsh ssh
    • Requires yubikey to be connected for re-login
    • Prompts for per-session MFA
  1. Upgrade auth settings to auth_service.authentication.require_session_mfa: hardware_key_touch
  • Using the existing login session (tsh ls) should force automatic re-login with yubikey
    • Prompts for touch if not cached
  • tsh ssh
    • Requires yubikey to be connected for re-login
    • Prompts for touch if not cached

Other @GavinFrazar

Set auth_service.authentication.require_session_mfa: hardware_key_touch in your cluster auth settings.

  • Database Acces: tsh proxy db
  • Application Access: tsh login app && tsh proxy app

Performance @rosstimothy @fspmarshall

Perform all tests on the following configurations:

  • With default networking configuration
  • With Proxy Peering Enabled
  • With TLS Routing Enabled
  • Cluster with 10K direct dial nodes:
  • etcd
  • DynamoDB
  • Firestore
  • Cluster with 10K reverse tunnel nodes:
  • etcd
  • DynamoDB
  • Firestore
  • Cluster with 500 trusted clusters:
  • etcd
  • DynamoDB
  • Firestore

Soak Test @rosstimothy @fspmarshall

Run 30 minute soak test with a mix of interactive/non-interactive sessions for both direct and reverse tunnel nodes:

tsh bench --duration=30m user@direct-dial-node ls
tsh bench -i --duration=30m user@direct-dial-node ps uax

tsh bench --duration=30m user@reverse-tunnel-node ls
tsh bench -i --duration=30m user@reverse-tunnel-node ps uax

Observe prometheus metrics for goroutines, open files, RAM, CPU, Timers and make sure there are no leaks

  • Verify that prometheus metrics are accurate.

Concurrent Session Test

  • Cluster with 1k reverse tunnel nodes

Run a concurrent session test that will spawn 5 interactive sessions per node in the cluster:

tsh bench sessions --max=5000 user ls
tsh bench sessions --max=5000 --web user ls 
  • Verify that all 5000 sessions are able to be established.
  • Verify that tsh and the web UI are still functional.

Robustness @rosstimothy @fspmarshall

  • Connectivity Issues:
  • Verify that a lack of connectivity to Auth does not prevent access to
    resources which do not require a moderated session and in async recording
    mode from an already issued certificate.
  • Verify that a lack of connectivity to Auth prevents access to resources
    which require a moderated session and in async recording mode from an already
    issued certificate.

Teleport with Cloud Providers @hugoShaka

AWS @hugoShaka

GCP @hugoShaka

  • Deploy Teleport to GCP. Using Cloud Firestore & Cloud Storage
  • Deploy Teleport to GKE. Google Kubernetes engine.
  • Deploy Teleport Enterprise to GCP.

IBM @hugoShaka

  • Deploy Teleport to IBM Cloud. Using IBM Database for etcd & IBM Object Store
  • Deploy Teleport to IBM Cloud Kubernetes.
  • Deploy Teleport Enterprise to IBM Cloud.

Application Access @mdwn

  • Run an application within local cluster.
    • Verify the debug application debug_app: true works.
    • Verify an application can be configured with command line flags.
    • Verify an application can be configured from file configuration.
    • Verify that applications are available at auto-generated addresses name.rootProxyPublicAddr and well as publicAddr.
  • Run an application within a trusted cluster.
    • Verify that applications are available at auto-generated addresses name.rootProxyPublicAddr.
  • Verify Audit Records.
    • app.session.start and app.session.chunk events are created in the Audit Log.
    • app.session.chunk points to a 5 minute session archive with multiple app.session.request events inside.
    • tsh play <chunk-id> can fetch and print a session chunk archive.
  • Verify JWT using verify-jwt.go.
  • Verify RBAC.
  • Verify CLI access with tsh app login.
  • Verify AWS console access.
    • Can log into AWS web console through the web UI.
    • Can interact with AWS using tsh aws commands.
  • Verify Azure CLI access with tsh app login.
    • Can interact with Azure using tsh az commands.
    • Can interact with Azure using a combination of tsh proxy az and az commands.
  • Verify GCP CLI access with tsh app login.
    • Can interact with GCP using tsh gcloud commands.
    • Can interact with Google Cloud Storage using tsh gsutil commands.
    • Can interact with GCP/GCS using a combination of tsh proxy gcloud and gcloud/gsutil commands.
  • Verify dynamic registration.
    • Can register a new app using tctl create.
    • Can update registered app using tctl create -f.
    • Can delete registered app using tctl rm.
  • Test Applications screen in the web UI (tab is located on left side nav on dashboard):
    • Verify that all apps registered are shown
    • Verify that clicking on the app icon takes you to another tab
    • Verify using the bash command produced from Add Application dialogue works (refresh app screen to see it registered)

Database Access @smallinsky

  • Connect to a database within a local cluster.
  • Connect to a database within a remote cluster via a trusted cluster.
  • Verify audit events. @GavinFrazar
    • db.session.start is emitted when you connect.
    • db.session.end is emitted when you disconnect.
    • db.session.query is emitted when you execute a SQL query.
  • Verify RBAC. @gabrielcorado
    • tsh db ls shows only databases matching role's db_labels.
    • Can only connect as users from db_users.
    • (Postgres only) Can only connect to databases from db_names.
      • db.session.start is emitted when connection attempt is denied.
    • (MongoDB only) Can only execute commands in databases from db_names.
      • db.session.query is emitted when command fails due to permissions.
    • Can configure per-session MFA. @GavinFrazar
      • MFA tap is required on each tsh db connect.
  • Verify dynamic registration. @GavinFrazar
    • Can register a new database using tctl create.
    • Can update registered database using tctl create -f.
    • Can delete registered database using tctl rm.
  • Verify discovery.
    • AWS @greedy52
      • Can detect and register RDS instances.
      • Can detect and register Aurora clusters, and their reader and custom endpoints.
      • Can detect and register RDS proxies, and their custom endpoints.
      • Can detect and register Redshift clusters.
      • Can detect and register Redshift serverless workgroups, and their VPC endpoints.
      • Can detect and register ElastiCache Redis clusters.
      • Can detect and register MemoryDB clusters.
    • Azure @GavinFrazar
      • Can detect and register MySQL and Postgres single-server instances.
      • Can detect and register MySQL and Postgres flexible-server instances.
      • Can detect and register Azure Cache for Redis servers.
      • Can detect and register Azure SQL Servers and Azure SQL Managed Instances.
  • Verify Teleport managed users (password rotation, auto 'auth' on connection, etc.). @greedy52
    • Can detect and manage ElastiCache users
    • Can detect and manage MemoryDB users
  • Test Databases screen in the web UI (tab is located on left side nav on dashboard): @smallinsky
    • Verify that all dbs registered are shown with correct name, description, type, and labels
    • Verify that clicking on a rows connect button renders a dialogue on manual instructions with Step 2 login value matching the rows name column
    • Verify searching for all columns in the search bar works
    • Verify you can sort by all columns except labels
  • Other
    • MySQL server version reported by Teleport is correct. @smallinsky

TLS Routing @smallinsky

  • Verify that teleport proxy v2 configuration starts only a single listener. @Tener
    version: v2
    teleport:
      proxy_service:
        enabled: "yes"
        public_addr: ['root.example.com']
        web_listen_addr: 0.0.0.0:3080
    
  • Run Teleport Proxy in multiplex mode auth_service.proxy_listener_mode: "multiplex" @Tener
    • Trusted cluster
      • Setup trusted clusters using single port setup web_proxy_addr == tunnel_addr
      kind: trusted_cluster
      spec:
        ...
        web_proxy_addr: root.example.com:443
        tunnel_addr: root.example.com:443
        ...
      
  • Database Access
  • Application Access @smallinsky
    • Verify app access through proxy running in multiplex mode
  • SSH Access @gabrielcorado
    • Connect to a OpenSSH server through a local ssh proxy ssh -o "ForwardAgent yes" -o "ProxyCommand tsh proxy ssh" [email protected]
    • Connect to a OpenSSH server on leaf-cluster through a local ssh proxyssh -o "ForwardAgent yes" -o "ProxyCommand tsh proxy ssh --user=%r --cluster=leaf-cluster %h:%p" [email protected]
    • Verify tsh ssh access through proxy running in multiplex mode
  • Kubernetes access: @Tener
    • Verify kubernetes access through proxy running in multiplex mode

Desktop Access @ibeckermayer

  • Direct mode (set listen_addr):
    • Can connect to desktop defined in static hosts section.
    • Can connect to desktop discovered via LDAP
  • IoT mode (reverse tunnel through proxy):
    • Can connect to desktop defined in static hosts section.
    • Can connect to desktop discovered via LDAP
  • Connect multiple windows_desktop_services to the same Teleport cluster,
    verify that connections to desktops on different AD domains works. (Attempt to
    connect several times to verify that you are routed to the correct
    windows_desktop_service)
  • Verify user input
    • Download Keyboard Key Info and
      verify all keys are processed correctly in each supported browser. Known
      issues: F11 cannot be captured by the browser without
      special configuration
      on MacOS.
    • Left click and right click register as Windows clicks. (Right click on
      the desktop should show a Windows menu, not a browser context menu)
    • Vertical and horizontal scroll work.
      Horizontal Scroll Test
  • Locking
    • Verify that placing a user lock terminates an active desktop session.
    • Verify that placing a desktop lock terminates an active desktop session.
    • Verify that placing a role lock terminates an active desktop session.
  • Labeling
    • Set client_idle_timeout to a small value and verify that idle sessions
      are terminated (the session should end and an audit event will confirm it
      was due to idle connection)
    • All desktops have teleport.dev/origin label.
    • Dynamic desktops have additional teleport.dev labels for OS, OS
      Version, DNS hostname.
    • Regexp-based host labeling applies across all desktops, regardless of
      origin.
  • RBAC
    • RBAC denies access to a Windows desktop due to labels
    • RBAC denies access to a Windows desktop with the wrong OS-login.
  • Clipboard Support
    • When a user has a role with clipboard sharing enabled and is using a chromium based browser
      • Going to a desktop when clipboard permissions are in "Ask" mode (aka "prompt") causes the browser to show a prompt when you first click or press a key
      • The clipboard icon is highlighted in the top bar
      • After allowing clipboard permission, copy text from local workstation, paste into remote desktop
      • After allowing clipboard permission, copy text from remote desktop, paste into local workstation
      • After disallowing clipboard permission, confirm copying text from local workstation and pasting into remote desktop doesn't work
      • After disallowing clipboard permission, confirm copying text from remote desktop and pasting into local workstation doesn't work
    • When a user has a role with clipboard sharing enabled and is not using a chromium based browser
      • The clipboard icon is not highlighted in the top bar and copy/paste does not work
    • When a user has a role with clipboard sharing disabled and is using a chromium and non-chromium based browser (confirm both)
      • The clipboard icon is not highlighted in the top bar and copy/paste does not work
  • Directory Sharing
    • On supported, non-chromium based browsers (Firefox/Safari)
      • Attempting to share directory logs a sensible warning in the warning dropdown
    • On supported, chromium based browsers (Chrome/Edge)
      • Begin sharing works
        • The shared directory icon in the top right of the screen is highlighted when directory sharing is initiated
        • The shared directory appears as a network drive named "<directory_name> on teleport"
        • The share directory menu option disappears from the menu
      • Navigation
        • The folders of the shared directory are navigable (move up and down the directory tree)
      • CRUD
        • A new text file can be created
        • The text file can be written to (saved)
        • The text file can be read (close it, check that it's saved on the local machine, then open it again on the remote)
        • The text file can be deleted
      • File/Folder movement
        • In to out (make at least one of these from a non-top-level-directory)
          • A file from inside the shared directory can be drag-and-dropped outside the shared directory
          • A folder from inside the shared directory can be drag-and-dropped outside the shared directory (and its contents retained)
          • A file from inside the shared directory can be cut-pasted outside the shared directory
          • A folder from inside the shared directory can be cut-pasted outside the shared directory
          • A file from inside the shared directory can be copy-pasted outside the shared directory
          • A folder from inside the shared directory can be copy-pasted outside the shared directory
        • Out to in (make at least one of these overwrite an existing file, and one go into a non-top-level directory)
          • A file from outside the shared directory can be drag-and-dropped into the shared directory
          • A folder from outside the shared directory can be drag-and-dropped into the shared directory (and its contents retained)
          • A file from outside the shared directory can be cut-pasted into the shared directory
          • A folder from outside the shared directory can be cut-pasted into the shared directory
          • A file from outside the shared directory can be copy-pasted into the shared directory
          • A folder from outside the shared directory can be copy-pasted into the shared directory
        • Within
          • A file from inside the shared directory cannot be drag-and-dropped to another folder inside the shared directory: a dismissible "Unsupported Action" dialog is shown
          • A folder from inside the shared directory cannot be drag-and-dropped to another folder inside the shared directory: a dismissible "Unsupported Action" dialog is shown
          • A file from inside the shared directory cannot be cut-pasted to another folder inside the shared directory: a dismissible "Unsupported Action" dialog is shown
          • A folder from inside the shared directory cannot be cut-pasted to another folder inside the shared directory: a dismissible "Unsupported Action" dialog is shown
          • A file from inside the shared directory can be copy-pasted to another folder inside the shared directory
          • A folder from inside the shared directory can be copy-pasted to another folder inside shared directory (and its contents retained)
    • RBAC
      • Give the user one role that explicitly disables directory sharing (desktop_directory_sharing: false) and confirm that the option to share a directory doesn't appear in the menu
  • Per-Session MFA (try webauthn on each of Chrome, Safari, and Firefox; u2f only works with Firefox)
    • Attempting to start a session no keys registered shows an error message
    • Attempting to start a session with a webauthn registered pops up the "Verify Your Identity" dialog
      • Hitting "Cancel" shows an error message
      • Hitting "Verify" causes your browser to prompt you for MFA
      • Cancelling that browser MFA prompt shows an error
      • Successful MFA verification allows you to connect
  • Session Recording
    • Verify sessions are not recorded if all of a user's roles disable recording
    • Verify sync recording (mode: node-sync or mode: proxy-sync)
    • Verify async recording (mode: node or mode: proxy)
    • Sessions show up in session recordings UI with desktop icon
    • Sessions can be played back, including play/pause functionality
    • Sessions playback speed can be toggled while its playing
    • Sessions playback speed can be toggled while its paused
    • A session that ends with a TDP error message can be played back, ends by displaying the error message,
      and the progress bar progresses to the end.
    • Attempting to play back a session that doesn't exist (i.e. by entering a non-existing session id in the url) shows
      a relevant error message.
    • RBAC for sessions: ensure users can only see their own recordings when
      using the RBAC rule from our
      docs
  • Audit Events (check these after performing the above tests)
    • windows.desktop.session.start (TDP00I) emitted on start
    • windows.desktop.session.start (TDP00W) emitted when session fails to
      start (due to RBAC, for example)
    • client.disconnect (T3006I) emitted when session is terminated by or fails
      to start due to lock
    • windows.desktop.session.end (TDP01I) emitted on end
    • desktop.clipboard.send (TDP02I) emitted for local copy -> remote
      paste
    • desktop.clipboard.receive (TDP03I) emitted for remote copy -> local
      paste
    • desktop.directory.share (TDP04I) emitted when Teleport starts sharing a directory
    • desktop.directory.read (TDP05I) emitted when a file is read over the shared directory
    • desktop.directory.write (TDP06I) emitted when a file is written to over the shared directory
  • Warnings/Errors
    • Induce the backend to send a TDP Notification of severity warning (1), confirm that a warning is logged in the warning dropdown
    • Induce the backend to send a TDP Notification of severity error (2), confirm that session is terminated and error popup is shown
    • Induce the backend to send a TDP Error, confirm that session is terminated and error popup is shown (confirms backwards compatibility w/ older w_d_s starting in Teleport 12)
  • Trusted Cluster / Tunneling
    • Set up Teleport in a trusted cluster configuration where the root and leaf cluster has a w_d_s connected via tunnel (w_d_s running as a separate process)
      • Confirm that windows desktop sessions can be made on root cluster
      • Confirm that windows desktop sessions can be made on leaf cluster

Binaries compatibility @tobiaszheller

  • Verify tsh runs on:
    • Windows 10
    • MacOS

Machine ID @timothyb89

SSH

With a default Teleport instance configured with a SSH node:

  • Verify you are able to create a new bot user with tctl bots add robot --roles=access. Follow the instructions provided in the output to start tbot
  • Verify you are able to connect to the SSH node using openssh with the generated ssh_config in the destination directory
  • Verify that after the renewal period (default 20m, but this can be reduced via configuration), that newly generated certificates are placed in the destination directory
  • Verify that sending both SIGUSR1 and SIGHUP to a running tbot process causes a renewal and new certificates to be generated
  • Verify that you are able to make a connection to the SSH node using the ssh_config provided by tbot after each phase of a manual CA rotation.

Ensure the above tests are completed for both:

  • Directly connecting to the auth server
  • Connecting to the auth server via the proxy reverse tunnel

DB Access

With a default Postgres DB instance, a Teleport instance configured with DB access and a bot user configured:

  • Verify you are able to connect to and interact with a database using tbot db while tbot start is running

Host users creation @lxea

Host users creation docs
Host users creation RFD

  • Verify host users creation functionality
    • non-existing users are created automatically
    • users are added to groups
      • non-existing configured groups are created
      • created users are added to the teleport-system group
    • users are cleaned up after their session ends
      • cleanup occurs if a program was left running after session ends
    • sudoers file creation is successful
      • Invalid sudoers files are not created
    • existing host users are not modified
    • setting disable_create_host_user: true stops user creation from occurring

CA rotations @espadolini

  • Verify the CA rotation functionality itself (by checking in the backend or with tctl get cert_authority)
    • standby phase: only active_keys, no additional_trusted_keys
    • init phase: active_keys and additional_trusted_keys
    • update_clients and update_servers phases: the certs from the init phase are swapped
    • standby phase: only the new certs remain in active_keys, nothing in additional_trusted_keys
    • rollback phase (second pass, after completing a regular rotation): same content as in the init phase
    • standby phase after rollback: same content as in the previous standby phase
  • Verify functionality in all phases (clients might have to log in again in lieu of waiting for credentials to expire between phases)
    • SSH session in tsh from a previous phase
    • SSH session in web UI from a previous phase
    • New SSH session with tsh
    • New SSH session with web UI
    • New SSH session in a child cluster on the same major version
    • New SSH session in a child cluster on the previous major version
    • New SSH session from a parent cluster
    • Application access through a browser
    • Application access through curl with tsh app login
    • kubectl get po after tsh kube login
    • Database access (no configuration change should be necessary if the database CA isn't rotated, other Teleport functionality should not be affected if only the database CA is rotated)

EC2 Discovery @lxea

EC2 Discovery docs

  • Verify EC2 instance discovery
    • Only EC2 instances matching given AWS tags have the installer executed on them
    • Only the IAM permissions mentioned in the discovery docs are required for operation
    • Custom scripts specified in different matchers are executed
    • Custom SSM documents specified in different matchers are executed
    • New EC2 instances with matching AWS tags are discovered and added to the teleport cluster
      • Large numbers of EC2 instances (51+) are all successfully added to the cluster
    • Nodes that have been discovered do not have the install script run on the node multiple times

Documentation @ptgott @alexfornuto

Checks should be performed on the version of documentation corresponding to the
major release we're testing for. For example, for Teleport 12 release use
branch/v12 branch and make sure to select "Version 12.0" in the documentation
version switcher.

Resources

Quick GitHub/SAML/OIDC Setup Tips

@russjones russjones added the test-plan A list of tasks required to ship a successful product release. label Jan 12, 2023
@Tener
Copy link
Contributor

Tener commented Jan 12, 2023

@mdwn I've added these points to the test plan, seemingly under your remit:

- [ ] Verify [Azure CLI access](https://goteleport.com/docs/ver/12.x/application-access/guides/azure/) with `tsh app login`.
  - [ ] Can interact with Azure using `tsh az` commands.
  - [ ] Can interact with Azure using a combination of `tsh proxy az` and `az` commands.
- [ ] Verify [GCP CLI access](https://github.com/gravitational/teleport/pull/19905) with `tsh app login`.
  - [ ] Can interact with GCP using `tsh gcloud` commands.
  - [ ] Can interact with Google Cloud Storage using `tsh gsutil` commands.
  - [ ] Can interact with GCP/GCS using a combination of `tsh proxy gcloud` and `gcloud`/`gsutil` commands.

Both Azure and GCP integrations are on master and will be part of the cut tomorrow. The PR for the GCP docs is in review #19905, but slightly out of sync with the implementation; this will be corrected early next week. Please don't hesitate to ask me for any clarifications or tips.

I'll send PR updating the test plan template too.

@GavinFrazar
Copy link
Contributor

I've added/edited these points to the test plan (for discovery and connect via local/remote cluster):

  - [ ] Azure single-server MySQL and Postgres
  - [ ] Azure flexible-server MySQL and Postgres

Added them to connect test because flexible server integration required an update to the way we modify db username in the engine.

Forgot to update the test plan template in #19759 I'll open a PR to update that template now as well.

@espadolini
Copy link
Contributor

Added "Changing role map of existing Trusted Cluster" here and in #20325

@tigrato
Copy link
Contributor

tigrato commented Jan 18, 2023

Added in #20274

@hugoShaka
Copy link
Contributor

hugoShaka commented Jan 18, 2023

tctl does not default to local auth: #20346

@timothyb89
Copy link
Contributor

timothyb89 commented Jan 18, 2023

Regression in tsh breaks identity file loading (affects most tbot proxying features, including ssh/db access): #20373

and another small issue where tbot's ssh_config forgets nonstandard ports: #20378

@GavinFrazar
Copy link
Contributor

#20384 issue with PIV yubikey integration

@mdwn
Copy link
Contributor

mdwn commented Jan 19, 2023

AWS console is inaccessible via the Teleport UI: #20385

@tigrato
Copy link
Contributor

tigrato commented Jan 19, 2023

The default access role misses permissions to list pods #20401

@codingllama
Copy link
Contributor

tsh login --auth=local uses platform passwordless if it can (#20429). Not a huge deal, as it does respect other settings/flags, but I'll take a look.

@mdwn
Copy link
Contributor

mdwn commented Jan 19, 2023

Setting Azure identities doesn't work for all valid characters in an identity string: #20434

@hugoShaka
Copy link
Contributor

Helm chart deadlock: #20488

@GavinFrazar
Copy link
Contributor

teleport db configure create --azure-sqlserver-discovery=$region generates invalid config yaml. fix here: #20496

@codingllama
Copy link
Contributor

I've found some issues with device trust, unusual verbs (create_enroll_token and enroll) and RoleAdmin. Pushing patches soon. (FYI @sfreiberg.)

@codingllama
Copy link
Contributor

I've found some issues with device trust, unusual verbs (create_enroll_token and enroll) and RoleAdmin. Pushing patches soon. (FYI @sfreiberg.)

Promised patches: #20505 and https://github.com/gravitational/teleport.e/pull/724. We'll need an e/ bump on branch/v12 after all is done.

@codingllama
Copy link
Contributor

tctl devices rm issue: #20506

@camscale
Copy link
Contributor

Okta SSO documentation setup issue: #20538

@rosstimothy
Copy link
Contributor

etcd Load Testing

Agent Mesh

10k Tunnel Nodes

image

https://teleportcoreteam.grafana.net/goto/9JtLQdTVz?orgId=1

10k Direct Dial Nodes

image

https://teleportcoreteam.grafana.net/goto/ss-CjOoVk?orgId=1

500 Trusted Cluster

image

https://teleportcoreteam.grafana.net/goto/yNS_PHo4z?orgId=1

Soak Test

----Direct Dial Node Test----
tsh --insecure --proxy=monster.gravitational.co:3080 -i /etc/teleport/auth bench --duration=30m root@node-6f44d86564-lqmrg ls

* Requests originated: 17999
* Requests failed: 0

Histogram

Percentile Response Duration
---------- -----------------
25         150 ms
50         152 ms
75         156 ms
90         160 ms
95         165 ms
99         191 ms
100        455 ms

----Reverse Tunnel Node Test----
tsh --insecure --proxy=monster.gravitational.co:3080 -i /etc/teleport/auth bench --duration=30m root@iot-node-86b9c86bff-fgrxs ls

* Requests originated: 17999
* Requests failed: 0

Histogram

Percentile Response Duration
---------- -----------------
25         146 ms
50         150 ms
75         154 ms
90         162 ms
95         170 ms
99         191 ms
100        411 ms

Proxy Peering

10k Tunnel Nodes

image

https://teleportcoreteam.grafana.net/goto/5i-Am-T4z?orgId=1

10k Direct Dial Nodes

image

https://teleportcoreteam.grafana.net/goto/cv55cFT4z?orgId=1

500 Trusted Cluster

image

https://teleportcoreteam.grafana.net/goto/Oog2Abo4z?orgId=1

Soak Test

----Direct Dial Node Test----
tsh --insecure --proxy=monster.gravitational.co:3080 -i /etc/teleport/auth bench --duration=30m root@node-6f44d86564-frcnx ls

* Requests originated: 17999
* Requests failed: 0

Histogram

Percentile Response Duration
---------- -----------------
25         147 ms
50         149 ms
75         153 ms
90         156 ms
95         160 ms
99         188 ms
100        352 ms

----Reverse Tunnel Node Test----
tsh --insecure --proxy=monster.gravitational.co:3080 -i /etc/teleport/auth bench --duration=30m root@iot-node-86b9c86bff-x4pbr ls

* Requests originated: 17999
* Requests failed: 0

Histogram

Percentile Response Duration
---------- -----------------
25         164 ms
50         169 ms
75         176 ms
90         186 ms
95         191 ms
99         209 ms
100        439 ms

@ravicious
Copy link
Member

ravicious commented Jan 24, 2023

teleport-ent macOS binaries are not signed. https://github.com/gravitational/teleport.e/issues/741

Edit: It turns out it's a known issue.

@codingllama
Copy link
Contributor

App Access and require_session_mfa issue: #20634.

@alexfornuto
Copy link
Contributor

alexfornuto commented Jan 24, 2023

Docker tctl command failure: #20637. Resolved

@capnspacehook
Copy link
Contributor

Error connecting to leaf OpenSSH node: #20703

@GavinFrazar
Copy link
Contributor

App Access and require_session_mfa issue: #20634.

Just want to note that my only remaining unchecked task is to use require_session_mfa: hardware_key_touch with app access, which also does not work (obviously)

@codingllama
Copy link
Contributor

App Access and require_session_mfa issue: #20634.

Just want to note that my only remaining unchecked task is to use require_session_mfa: hardware_key_touch with app access, which also does not work (obviously)

Isn't hardware_key_touch functionally equivalent to no session MFA, due to the checks being client-side? Maybe chat with @Joerger to figure out if it's actually supposed to work.

@Joerger
Copy link
Contributor

Joerger commented Jan 26, 2023

Just want to note that my only remaining unchecked task is to use require_session_mfa: hardware_key_touch with app access, which also does not work (obviously)

Isn't hardware_key_touch functionally equivalent to no session MFA, due to the checks being client-side? Maybe chat with @Joerger to figure out if it's actually supposed to work.

Yes, hardware_key_touch does actually work in v12, but hardware_key does not as it is functionally equivalent to require_session_mfa: yes. I've tested it so I'll check it off on the test plan.

Edit: I was wrong on this and the test I performed was inadequate (tsh proxy app does prompt for tap, but using the connection fails due to the app session using a different key on the server). This test should have been removed before as this lack of support was already discovered and documented.

@GavinFrazar
Copy link
Contributor

tsh proxy aws --endpoint-url not working in alpha.1 & alpha.2: #20798

I think this broke when I did some refactoring work in app access.

@GavinFrazar
Copy link
Contributor

tsh db connect fails when PIV is enabled (not in all configurations): #20799

I didn't notice this in my first pass through the test plan because I didn't trip on one of the configurations that has the issue.

@fspmarshall
Copy link
Contributor

DynamoDB

Direct Dial Scaling

10k-direct-scaling

Direct Dial Soak

$ tsh bench --duration=30m root@ip-172-31-8-224-us-west-2-compute-internal ls

* Requests originated: 17999
* Requests failed: 0

Histogram

Percentile Response Duration
---------- -----------------
25         156 ms
50         165 ms
75         178 ms
90         188 ms
95         194 ms
99         219 ms
100        2767 ms
$ tsh bench --interactive --duration=30m root@ip-172-31-8-224-us-west-2-compute-internal ps aux

* Requests originated: 17999
* Requests failed: 0

Histogram

Percentile Response Duration
---------- -----------------
25         164 ms
50         174 ms
75         186 ms
90         194 ms
95         201 ms
99         223 ms
100        1677 ms

Tunnel Scaling

10k-tunnel-scaling

Tunnel Soak

$ tsh bench --duration=30m root@ip-172-31-8-224-us-west-2-compute-internal ls


* Requests originated: 17999
* Requests failed: 0

Histogram

Percentile Response Duration
---------- -----------------
25         176 ms
50         183 ms
75         192 ms
90         204 ms
95         212 ms
99         250 ms
100        2229 ms
$ tsh bench --interactive --duration=30m root@ip-172-31-8-224-us-west-2-compute-internal ps aux


* Requests originated: 17998
* Requests failed: 0

Histogram

Percentile Response Duration
---------- -----------------
25         184 ms
50         191 ms
75         201 ms
90         213 ms
95         220 ms
99         260 ms
100        2805 ms

@fspmarshall
Copy link
Contributor

Note about ssh agent forwarding and ssh file copying RBAC tests: The RBAC section of the testplan mentions that we expect access denied to show up in the audit log for all items, but ssh agent forwarding and ssh file copying do not generate access denied events. I checked the code for these checks, and it appears that they aren't intended to emit events currently, so this doesn't seem to be a regression.

I've marked these sections as complete in the testplan because teleport seems to be working as intended, but it may be worth considering adding access denied events for these items.

@alexfornuto
Copy link
Contributor

Verify Teleport versions throughout documentation are correct and reflect upcoming release:

#20643

@hugoShaka
Copy link
Contributor

hugoShaka commented Jan 30, 2023

teleport-cluster chart v12 breaks when used with etcd backend: #20960

@tigrato
Copy link
Contributor

tigrato commented Jan 31, 2023

Pod RBAC fails if the Kubernetes Vendor runs with compression enabled: #20980 - PR: #20981

@hugoShaka detected this issue when running a test in the IBM cloud

@ibeckermayer
Copy link
Contributor

Issue and fix here: #21009

@alexfornuto
Copy link
Contributor

Verify upcoming releases page is accurate:

This is under Documentation, but we don't have insight into what's being slated for the next release. I've prepped a page in #21283 for data to be added to, but maybe this should be considered part of #21317

@zmb3 zmb3 closed this as completed Feb 9, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
test-plan A list of tasks required to ship a successful product release.
Projects
None yet
Development

No branches or pull requests