Skip to content

Latest commit

 

History

History
165 lines (117 loc) · 5.26 KB

README.md

File metadata and controls

165 lines (117 loc) · 5.26 KB

PLSC, a Python implementation of LSC

This project is a Python implementation aiming for LDAP to LDAP synchronization. We are inspired by the LSC project (https://lsc-project.org/doku.php)

Install

For local development we advice to make use of Python Virtual Environment. The installation commands are specified below.

Sample Synchronisation configuration

This is an example of what we need to specify source and destination. [ More complex examples will follow ]

---
ldap:
  src:
    uri: ldap://ldap.vm.scz-vm.net
    basedn: dc=scz,dc=vnet
    binddn: cn=John Doe,dc=scz,dc=vnet
    passwd: changethispassword
  dst:
    uri: ldap://ldap.vm.scz-vm.net
    basedn: dc=clients,dc=vnet
    binddn: cn=admin,dc=clients,dc=vnet
    passwd: changethispassword
    sbs:
src:
  host: https://sbs.example.net
  user: sysread
  passwd: changethispassword

Local development

Create a virtual environment and install the required python packages:

python3 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install -r requirements.txt

Use

sync_services will look for all labeledURI attributes in the source and create dc=<labeledURI> branches on the destination containing all CO's that are linked to these services with bare People, Groups and uid/gidNumberSequence structures.

plsc will then mirror all People and Groups to the corresponding CO's in the destination LDAP, optionally converting attributes and dn's as defined in the code on the way:

            # Here's the magic: Build the new person entry
            dst_entry = {}
            dst_entry['objectClass'] = ['inetOrgPerson', 'person', 'posixAccount']
            dst_entry['uid'] = [src_uid]
            dst_entry['cn'] = src_entry['cn']
            dst_entry['sn'] = src_entry['sn']
            dst_entry['homeDirectory'] = ['/home/{}'.format(src_uid)]

And for groups:

            # Here's the magic: Build the new group entry
            m = re.search('^(?:GRP)?(?:CO)?(?:COU)?:(.*?)$', src_cn)
            dst_cn = src_type + "_" + m.group(1) if m.group(1) else ""

            dst_entry = {}
            dst_entry['objectClass'] = ['extensibleObject', 'posixGroup', 'sczGroup']
            dst_entry['cn'] = [dst_cn]
            dst_entry['description'] = [src_cn]

Testing

For local testing, you need a local LDAP to be started before running tests. When you have docker installed on your local machine, you can simple run:

etc/ldap_start.sh

This script will take care of pulling the proper LDAP image, start the docker container and initialize it properly with all required objectclasses.

If you do not have docker installed or wish to use an existing running LDAP server, then please make sure that the config files listed in etc/ldif are properly installed.

You can specify LDAP connection and access constants in a local .env file, for exanple:

LDAP_URL="ldap://localhost:389"
LDAP_ADMIN_PASSWORD="secret"
LDAP_CONFIG_PASSWORD="config"
LDAP_DOMAIN="sram.tld"
LDAP_BASE_DN="dc=sram,dc=tld"
LDAP_BIND_DN="cn=admin,dc=sram,dc=tld"

SBS_URL=https://sbs.example.com
SBS_USER=sysread
SBS_PASS=secret
SBS_API_RECORDING=Yes
SBS_VERIFY_SSL=Yes

You have the option to run against an operational instance of SBS by specifing the SBS_URL and SBS_USER /SBS_PASS constants as shown above. If you do not want to access an SBS instance, just leave these constant out. In that case the SBS API's are immitated by the results listed in the local api directory of this repository.

In case you are testing against an operational SBS instance, you have the option to record the API results for later use during mockup testing, just set the environment variable SBS_API_RECORDING to "Yes". Now the API requests results will be stored in the local directory under ./api/...

When you omit the SBS_URL variable, the tests will run API requests agains the contents off this local ./api/... directory

When all these preperations are completed, you can now run the tests:

pytest -v

After each Test cycle the contents of the LDAP can be inspected, for that run this command:

etc/ldap_show.sh

When finished you can tear down the local running LDAP instance, by:

etc/ldap_stop.sh

Using Make

A Makefile is added to make life easier. Now you can run:

make pytest

Or just make. This will build the docker image and run pytest next.

make

Docker

A Dockerfile is added to produce an image that is capable of running PLSC. Since PLSC depends on an input YAML file holding the source and destionatination details, you have to provide sucan YANL file. Using Docker you can do the following:

Suppose you have prepared you YAML file (inspired by plsc.yml.example) and this file is 'my-plsc.yml'

You have to mount this file into the container and then run run.sh with that file as parameter.

Example:

docker run -v ${PWD}:/opt/plsc plsc ./run.sh my-plsc.yaml

If you started the local LDAP via etc/ldap_start.sh then you can connect to that if you run this PLSC container in network = host mode.

docker run -v ${PWD}:/opt/plsc --network host plsc ./run.sh my-plsc.yaml

run.sh is the existing script that runs both plsc_ordered.py and plsc_flat