Multi tenant enterprise grade IAM implementation based on a geo-distributed layered Vault installation. Currently under active development of initial version.

needs:

1. linux or macos system (only intel processors, for apple silicone using needs some changes in scripts)
2. docker

./build.sh

possible options for build:

./build.sh plugins # build separate plugins

./build.sh authd  # builds authd component

./build.sh cli # builds cli utility

./build.sh server-accessd # builds server-accessd component

./build.sh nss # builds nss component

./build.sh oidc-mock # builds oidc-mock for e2e tests purposes only 

./build.sh vault          # builds complete vault with plugins onboard
./build.sh vault --force  # builds complete vault with plugins onboard (use after first build)  

There are three possible modes of test and stage environment:

1. One vault in dev mode, negentropy plugins are aside (SINGLE mode)
2. Vaults with negentropy plugins onbоard (E2E mode, for E2E Tests in CI)
3. Vaults with negentropy plugins onbоard, run under delve debugger (DEBUG mode)

./start.sh single

runs one vault at docker container, uses separate plugin binaries, placed at vault-plugins/build

./start.sh e2e

runs several vaults at docker-containers, uses complete vault binary with negentropy plugins onboard, placed at infra/common/vault/vault/bin

./start.sh debug

runs several vaults at docker-containers, each docker run under delve debugger server uses complete vault binary with negentropy plugins onboard, placed at infra/common/vault/vault/bin, need connection delve-client debuggers to localhost: 2345 and localhost:2346 (see docker/docker-compose.debug.yml)

1. Zookepper, Kafka used to save data and communicate plugins.
2. Kafdrop used to study Kafka
3. test-server used as a sample of server under negentropy access control
4. test-client used as a sample of user PC, accessing servers under negentropy access control
5. oidc-mock provide mock of oidc-provider for tests

1. run all components containers
2. configure negentropy plugins
3. export data for running tests and unsealing vaults

./run-e2e-tests.sh 

1. No panic which can run at vault-plugins except:

   • panic run (or not) depends on code compositions only
   • panic run (or not) in tests runs
   • panic at flant-gitops plugin
     Check there is no panic with comment '// nolint:check_panic' at others places

2. Each new category stored in memdb should be mentioned at:

   • memdb schema
   • ~kafka_destination/vault.go isValidObjectType()` func
   • ~kafka_destination/metadata.go isValidObjectType() func
   • ~kafka_source/self.go Restore() func (or ../root.go)
   • checking of normal saving/restoration at e2e/tests/restoration/all_restoration_test.go

1. Set configuration for your namespace in .helm/values.yaml (domain, balancer, bucket if needed)

Example:

load_balancer:
  mynamespace: enabled
domain:
  mynamespace: mynamespace-negentropy.flant.dev

2. Go to teleport https://teleport.negentropy.flant.dev/ and connect to negentropy server console
3. Set target namespace for negentropy (required)

Example:

export NS=mynamespace

4. Set git branch for watching (required)

Example:

export GIT_BRANCH=myBranch

5. Set INITIAL_LAST_SUCCESSFULL_COMMIT (required)

Example:

export INITIAL_LAST_SUCCESSFULL_COMMIT=49ddefbced3654c669f0d139be4690477d9bd5aa

6. Set optional params GIT_REPO, REQUIRED_NUMBER_OF_SIGNATURES, GIT_POLL_PERIOD

export REQUIRED_NUMBER_OF_SIGNATURES=3 # default 0
export GIT_REPO=https://github.com/mygithub/negentropy.git # default https://github.com/flant/negentropy.git
export GIT_POLL_PERIOD=30m # default 1m

7. Run bootstrap script

curl -s https://raw.githubusercontent.com/flant/negentropy/$GIT_BRANCH/bootstrap-kube.sh| bash