Design Objectives
The design objective of any SANsymphony configuration is an ‘end-to-end’ process (from the user to the data) as well as availability, redundancy and performance. None of these can be considered in isolation but as integrated parts of an entire infrastructure. The information in this document provides some high-level recommendations that are often overlooked or simply forgotten about.
Avoid Complexity
The more complex the design, the more likely unforeseen problems will occur. A complex design can also make a system difficult to maintain and support, especially as it grows. A simple approach is recommended whenever possible.
Avoid Single Points of Failure
Dependencies between major components (e.g., switches, networks, fabrics and storage arrays) can impact the entire environment if they fail.
- Distribute each component over separate IO boards, racks, separate rooms, separate buildings and even separate 'sites'.
- Keep storage components away from public networks.
- Avoid connecting redundant devices to the same power circuit.
- Use redundant UPS-protected power sources and connect every device to it. A UPS backup does not help much if it fails to notify a Host to shut down because that Host’s management LAN switch itself was not connected to the UPS-backed power circuit.
- Use redundant network infrastructures and protocols where the failure of one component does not make access unavailable to others.
- Do not forget environmental components. A failed air conditioner may collapse all redundant systems located in the same data center. Rooms on the same floor may be affected by a single burst water pipe (even though they are technically separated from each other).
- DataCore strongly recommends that third-party recommendations are followed, and that all components (hardware and software) are in active support with their respective vendors.
Set Up Monitoring and Event Notification
Highly available systems keep services alive even if half of the environment has failed but these situations must always be recognized and responded to promptly so that they can be fixed as soon as possible by the personnel responsible to avoid further problems.
Documentation
Document the environment properly, keeping it up-to-date and accessible. Establish 'shared knowledge' between at least two people who have been trained and are familiar with all areas of the infrastructure.
User Access
Make sure that the difference between a 'normal' server and a ‘DataCore Server’ is understood. A DataCore Server should only be operated by a trained technician.
Best Practices are Not Pre-requisites
Some of the Best Practices listed here may not apply to your installation or cannot be applied because of the physical limitations of your installation – for example, it may not be possible to install any more CPUs or perhaps add more PCIe Adaptors, or maybe your network infrastructure limits the number of connections possible between DataCore Servers and so on.
Therefore, each set of recommendations is accompanied by a more detailed explanation so that when there is a situation where a best practice cannot be followed, there is an understanding of what and how that may limit your SANsymphony installation.
Also see:
How to become a DataCore Certified Implementation Engineer (DCIE)