compliance

package
v0.0.0-...-43a96cf Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Dec 19, 2024 License: Apache-2.0 Imports: 60 Imported by: 1

README

Compliance Service

Getting Started

  • Our all-in-one service (the same one we ship):
    run make run-with-es-pg from here (components/compliance-service)

This all-in-one service includes reporting, scanner, and inspec-agent, which can each be run individually is so desired....but we recommend just running the whole shabang. Making great things here people, don't want you to miss out on alllll the compliance amazingness!!

Notes:

  • Running the all-in-one service will start grpc on port 10121

    Something gone wrong? Take a look in /tmp/compliance.log to view the logs for the service.

  • Should you wish to view the log output live in your terminal, you can instead run make start-es-pg && make run-with-logs. This is useful when debugging.

Development Information

Development

Make sure you have Go installed! https://golang.org/dl/

Make some edits and run the binary with the commands provided in the Makefile. Tada!

Dependency Management

We use Go's module system for dependency management. See https://blog.golang.org/using-go-modules for an overview.

Sample Data

To load sample data for the reporting (ElasticSearch based service): make ingest-reports-into-es _please note: this will send in three reports from 2017. You will need to adjust your calendar section in the ui to view this data_

To load sample market profiles: make download-sample-market-profiles

To send a sample report to the ingest service: make send-ingest-report

Three nodes will be ingested:

  • "debian(1)-zeta-linux(f)-apache(s)-failed" - scan: 2018-02-09
  • "windows(1)-zeta-apache(s)-skipped" - scan: 2018-03-04
  • "centos(2)-beta-nginx(p)-apache(s)-passed" - scan: 2018-03-04, 2018-03-05

following this node name convention:

platform(number_of_total_scans)-environment-profile(status)-profile(status)-overall_status

Use the scanner tests to load a bunch of data, and reports generated from jobs:
make start-ssh-node && make test-scanner NO_DELETE="true" <- this will give you now-time scan reports.

Login to your cc_pg (postgres container): make login-pg Clear your postgres data: make clear-pg Clear your elastic data: make clear-es

Testing

The integration tests for the compliance service are written using ruby minitest. All tests and test-support files can be found in compliance-service/api/tests.

What do we run in CI? make test-integration-reporting, make test-db, make test-integration-scanner, as well as the unit tests, make fmt and make test-unit

To run the smoke tests, inspec exec inspec/a2-api-integration --sudo --target ssh://USERNAME@a2-local-inplace-upgrade-dev.cd.chef.co -i KEY from root of a2, using inspec version ~1.51

To run reporting, scanner, or nodemanager tests,

  • make kill-all
  • make test-prep
  • make run-with-logs
  • in a new window, make test-reporting, make test-scanner, or make test-nodemanager.

Easiest way to run the tests: make run-test. This will also do all the setup stuff/run the service for you. Great for set and forget *Easiest way to run partial tests: TEST="04_stats*" make run-test

Some notes:

  • make kill-all is a good thing to run first, all the time
  • make test-prep will get all the things ready; it will run:
    • make download-sample-market-profiles to get the sample profiles needed for testing
    • make start-ssh-node to start the ssh node we test against in 40_jobs_ssh_spec
    • gem install inspec to get latest inspec.
  • make test-reporting will always start by ingesting some reports. You will see the service re-index afterwards. This is normal.
  • make test-scanner will test scanner including scheduler and ingest connection.
  • make test-nodemanager will run the ingest to manager connection test and then all the nodemanager tests.

Test ElasticSearch migrations: A1_DATA=true TEST="60_A1_migration_spec.rb" make test-automate-upgrade A2V1_DATA=true TEST="61_A2V1_migration_spec.rb" make test-automate-upgrade A2V2_DATA=true TEST="62_A2V2_migration_spec.rb" make test-automate-upgrade A2V3_DATA=true TEST="63_A2V3_migration_spec.rb" make test-automate-upgrade A2V4_DATA=true TEST="64_A2V4_migration_spec.rb" make test-automate-upgrade A2V5_DATA=true TEST="65_A2V5_migration_spec.rb" make test-automate-upgrade A2V6_DATA=true TEST="66_A2V6_migration_spec.rb" make test-automate-upgrade

When in doubt, run the same commands we run in CI

You will also need the following two env variables:

  • AWS_ACCESS_KEY_ID
  • AWS_SECRET_ACCESS_KEY to run the aws-related-tests. You may already have IAM creds in ~/.aws/credentials, or you can create new IAM creds from the AWS UI.

To run the azure-related tests, you'll need the following env vars:

  • AZURE_CLIENT_ID
  • AZURE_CLIENT_SECRET
  • AZURE_TENANT_ID These credentials can be found in LastPass under "inspec azure test creds". The azure tests will automatically get skipped if you do not have these env vars set.

When the test command is run, the ruby grpc code is generated and the GRPC function is called against that with the request (see https://github.com/chef/automate/components/compliance-service/blob/master/api/tests/run.rb#L316.

Note: if you ever run into a problem while running the tests (or you simply need to debug a test you're writing), require 'pry';binding.pry is your best friend! Stick that right before the expectation that is failing to fall into a pry breakpoint and see what's going on.

Note: trouble running the tests because the ruby grpc bits are bombing out? run go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-grpc-gateway

The unit tests are run by executing make test-unit-reporting and make-test-unit-scanner. The db tests are run by executing make test-db.

There are many other wonderful testing targets, please take a look at testing section of the Makefile to see them all!

GRPC/Protobuf

Whenever you make a change to a *.proto file, you'll need to run make proto from the root of this repo to generate the updated code.

DB (Postgres)

We are using https://github.com/golang-migrate/migrate for migration. It happens in-code, triggered from the dao/pgdb/db.go. The code for the migration bits is in lib/db/migrator. To make a change to the schema, create an up and down file in dao/pgdb/migration/sql following the established pattern.

Repo structure and services information

Repo Structure

Here's a quick overview of what's going on in this repo:

  • Start with cmd/compliance-service/main.go. That calls cmd.RootCmd.Execute(), which initializes our cobra cli app (cmd/compliance-service/cmd/run.go).

  • The cobra cli app sets up all the flags we need to run the app. It initializes our db connection to postgres, and calls out to automate.Serve (in the root of the repo, compliance.go) with the db instance, configuration options, and grpc binding info.

  • The content in server.go sets the es backend and initializes our grpc routes. All grpc routes can be found in compliance-service/api.

Services Information

  • reporting: what is it and why?

    Reporting is our elasticsearch based service. It reports on the state of your infrastructure based on the InSpec reports sent in (for automate 1.x, that's via data-collector, for automate 2.0, that's via ingest-service). This includes api/stats, api/reporting, api/profiles.

  • scanner: what is it and why?

    Scanner is our postgres based service. It provides the user with the ability to schedule InSpec jobs on specified nodes. This includes api/secrets, api/nodes, api/jobs. The expected user flow is: user adds a secret, user adds a node (with a secret id for reference), user adds a job, specifying the desired nodes and profiles to be used for that job. That job information is then sent over to inspec-agent for execution.

  • manager: what is it and why?

    Please refer to https://github.com/chef/automate/blob/master/components/compliance-service/manager/README.md for more information.

  • inspec-agent: what is it and why?

    The inspec-agent takes job information and executes it. It initializes a worker pool based on the provided values for JobBufferSize (max num of jobs that can be accepted in worker queue) and JobWorkers (max num of workers to run in parallel). The service exposes a jobs endpoint for retrieving information on job status, and a workers endpoint for information on worker status, i.e. ./inspec-agent workers.

A2 Build

Testing things in the whole A2 Environment

From the root of A2, hab studio enter. Run start_deployment_service, followed by chef-automate dev deployinate. Run build components/compliance-service to get a local build of the compliance service. Anything not explicitly built will be pulled from the depot. Run generate_supertoken to get a token for curling the api.

Documentation

Index

Constants

This section is empty.

Variables

View Source 
var (
	PurgeWorkflowName = cereal.NewWorkflowName("purge")
	PurgeScheduleName = "periodic_purge"
)
View Source 
var SERVICE_STATE serviceState

Functions

func Serve 

func Serve(conf config.Compliance, grpcBinding string) error

Serve grpc

Types

type NodeManagerMock 

type NodeManagerMock struct {
}

func (*NodeManagerMock) ChangeNodeState 

func (nm *NodeManagerMock) ChangeNodeState(ctx context.Context, in *manager.NodeState,
	opts ...grpc.CallOption) (*manager.ChangeNodeStateResponse, error)

func (*NodeManagerMock) Connect 

func (nm *NodeManagerMock) Connect(ctx context.Context, in *manager.NodeManager,
	opts ...grpc.CallOption) (*empty.Empty, error)

func (*NodeManagerMock) ConnectManager 

func (nm *NodeManagerMock) ConnectManager(ctx context.Context, in *manager.Id,
	opts ...grpc.CallOption) (*empty.Empty, error)

func (*NodeManagerMock) Create 

func (nm *NodeManagerMock) Create(ctx context.Context, in *manager.NodeManager,
	opts ...grpc.CallOption) (*manager.Ids, error)

func (*NodeManagerMock) Delete 

func (nm *NodeManagerMock) Delete(ctx context.Context, in *manager.Id,
	opts ...grpc.CallOption) (*empty.Empty, error)

func (*NodeManagerMock) DeleteWithNodeStateStopped 

func (nm *NodeManagerMock) DeleteWithNodeStateStopped(ctx context.Context, in *manager.Id,
	opts ...grpc.CallOption) (*empty.Empty, error)

func (*NodeManagerMock) DeleteWithNodeStateTerminated 

func (nm *NodeManagerMock) DeleteWithNodeStateTerminated(ctx context.Context, in *manager.Id,
	opts ...grpc.CallOption) (*empty.Empty, error)

func (*NodeManagerMock) DeleteWithNodes 

func (nm *NodeManagerMock) DeleteWithNodes(ctx context.Context, in *manager.Id,
	opts ...grpc.CallOption) (*manager.Ids, error)

func (*NodeManagerMock) GetNodeWithSecrets 

func (nm *NodeManagerMock) GetNodeWithSecrets(ctx context.Context, in *manager.Id,
	opts ...grpc.CallOption) (*nodes.Node, error)

func (*NodeManagerMock) List 

func (nm *NodeManagerMock) List(ctx context.Context, in *manager.Query,
	opts ...grpc.CallOption) (*manager.NodeManagers, error)

func (*NodeManagerMock) ProcessNode 

func (nm *NodeManagerMock) ProcessNode(ctx context.Context, in *manager.NodeMetadata,
	opts ...grpc.CallOption) (*manager.ProcessNodeResponse, error)

func (*NodeManagerMock) Read 

func (nm *NodeManagerMock) Read(ctx context.Context, in *manager.Id,
	opts ...grpc.CallOption) (*manager.NodeManager, error)

func (*NodeManagerMock) SearchManagerNodes 

func (nm *NodeManagerMock) SearchManagerNodes(ctx context.Context, in *manager.NodeQuery,
	opts ...grpc.CallOption) (*manager.ManagerNodes, error)

func (*NodeManagerMock) SearchNodeFields 

func (nm *NodeManagerMock) SearchNodeFields(ctx context.Context, in *manager.FieldQuery,
	opts ...grpc.CallOption) (*manager.Fields, error)

func (*NodeManagerMock) SearchNodes 

func (nm *NodeManagerMock) SearchNodes(ctx context.Context, in *manager.NodeQuery,
	opts ...grpc.CallOption) (*manager.Nodes, error)

func (*NodeManagerMock) Update 

func (nm *NodeManagerMock) Update(ctx context.Context, in *manager.NodeManager,
	opts ...grpc.CallOption) (*empty.Empty, error)

type NotifierMock 

type NotifierMock struct {
}

func (*NotifierMock) QueueSize 

func (n *NotifierMock) QueueSize() int

func (*NotifierMock) Send 

func (n *NotifierMock) Send(context.Context, *notifications.Event)

type ServiceInfo 

type ServiceInfo struct {
	HostBind   string
	Port       int
	ServerBind string
	// contains filtered or unexported fields
}

ServiceInfo holds service listen info

func NewServiceConfig 

func NewServiceConfig(cfg *config.Compliance, connFactory *secureconn.Factory) *ServiceInfo

NewServiceConfig returns a ServiceInfo instance

func (*ServiceInfo) ProfileTarHandler 

func (conf *ServiceInfo) ProfileTarHandler(w http.ResponseWriter, r *http.Request)

ProfileTarHandler is the http handler for profile tarballs, used by the inspec-agent for executing scans

Source Files

View all Source files

Directories

Path Synopsis
api
jobs/server
profiles/conversion
profiles/server
reporting/server
stats/server
status/server
version/server
cmd
compliance-service
compliance-service/cmd
inspec_runner
inspec_runner/platform
dao
pgdb
pgdb/dbtest
testDB
examples
helpers
ingest
ingest/test
profiles
profiles/test
reporting
reporting/test
ingest
events/compliance
ingestic
ingestic/mappings
pipeline
pipeline/message
pipeline/processor
pipeline/publisher
server
inspec-agent
remote
resolver
runner
scheduler
types
profiles
db
market
relaxting
util
logging
pool

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.