When building the IT room and facilities the following aspects should be taken into account:
- Build and / or upgrade and test the IT room and all associated facilities such as UPS, HVAC, racking, cabling etc.,
- Build and test hardware infrastructure using the macro and micro designs,
- Build and / or extend existing and test all required shared services and Disaster Recovery using the macro and micro designs, and
- Build and test the operating system, middleware and application stacks on the base hardware infrastructure.
Today we are living in a free and open society where all details for building IT room architectures can be found on the open Internet. The giant IT companies like Google, Facebook and Amazon are all sharing their architectures with the open world. Just open your browser and search on ‘build datacenter Google’, ‘architecture datacenter Google’, etc. for the details of their datacenters. Actually, everyone can be part of the Facebook community for designing their datacenters. They have a project initiated where everyone can join voluntarily to become part of discussing and contributing to the Facebook datacenter’s. This project is called ‘Open Compute Project’. Hereby some links where to find more information:
To build the IT room it’s recommended to have this performed by a specialized company. The macro and micro designs should be followed and adhered to. Keep this documentation updated when changes will occur for solid reference in the future.
Tasks
- Interlock with ‘identify long lead time’ elements.
- Engage with IT room facilities management team.
- Finalise all target IT room building works.
- Provision and test appropriate electrical power, air conditioning, UPS and structured cabling.
- Establish and implement appropriate physical security and procedures.
- Obtain rack layouts from IT room facilities management team and verify adequacy.
- Build and/or extend shared services based on macro and micro designs.
- Test shared services.
- Complete target design and support documentation for each shared services.
- Finalise hardware installations.
- Configure infrastructure based on macro and micro designs.
- Perform infrastructure verification tests.
- Complete target design and support documentation for the target infrastructure.
- Build and/or extend disaster recovery based on macro and micro designs.
- Test disaster recovery capability.
- Complete target design and support documentation for new disaster recovery service.
- Install, configure and test base operating system and all appropriate patches.
- Install, configure and test middleware stack.
- Install, configure and test application code and all appropriate patches.
- Perform integration tests.
- Complete target design and support documentation for the target infrastructure.
- Update DCM Data Repository
Hints and tips
- Identify the organisation responsible for ‘racking and stacking’ equipment and for space and power.
- Ensure that built shared services, servers, disaster recovery are deployed into a ‘safe’ zone to enable safe proving and testing.
- Exercise great care if live source shared services are ‘extended’ to the target IT room. Safeguard source systems as a top priority!
- Understand the technical and operational mitigations if source shared services, servers, and disaster recovery are inadvertently affected.
- If building in a ‘live’ zone, understand the technical and operational mitigations if source servers are inadvertently affected.
- Continually monitor source service for potential service impacts in source IT room.
- Engage the service delivery organisation at this stage to ensure they are involved and committed to support the target IT room deployment when it eventually goes ‘live’.
- Exercise great care if changing the existing disaster recovery service as it needs to support all production services during the various stages of the migration.
- Ensure that the necessary security controls are in place before connecting source and target IT room together and obtain the necessary security approvals to make such connections
- Beware of, and monitor, any data replication impacts on live service (I/O rates, CPU demand), use ‘throttling’ if necessary.
- Ensure that build conform to build instructions and that any deviations are captured and the source build documentation updated.
Activity output
- IT room physical layout diagram
- Target IT room physical provision completion report
- Shared services test completion report
- Target infrastructure build completion report
- Target infrastructure test completion report
- Disaster recovery test completion report
- Test and usability reports
- Target OS, middleware and application build completion report
- Target OS, middleware and application test completion report
About the Author
