2. SCALABLE SAN IMPLEMENTATION PLAN CREATION
2.1 Verify and plan for dual power feeds for all components.
- DUAL is the key word
- Dual Feeds
- Dual PDUs
- Dual Power Supplies
- Controllers and shelves all have dual power supplies. Power should be plugged into A and B circuits. Each leg of power should be able to handle the load of the entire environment (N+N)
- Refer to Site Requirements Guide for more information, and power/temp specs.
- Verify that Controllers/Shelves were ordered with the proper plug style for your nation/voltage/outlet. (For example NEMA L5-20 vs IEC c13-c14. See common cable types here. )
2.2 Be able to create cabinet diagrams or be able to read and interpret a cabinet diagram. Diagrams should include the cabinet’s storage systems and switches with all connections shown.
- Core info
- Rack Location
- Cabinet U positions of servers
- Switch Placement
- Utilize Visio or NetApp Synergy to mock of the layout and available space.
- If they show exact ports to exact switches, keep an eye on what may be trunked, that FC isnt plugged into the wrong ports, etc.
2.3 Create a connectivity diagram. Be able to read and interpret a connectivity diagram.
- Fibre Channel
- Port names/Locations
- Initiator vs Target port setting
- FC vs UTA
- FC Switch port connectivity
- WWPN
- Port Locations
- Host name, Zone name, Aliases
- IP
- Switch name / port
- VLANs
- Trunks / LACP
- VIF names
- See section 3 for more detailed information
2.4 Plan storage controller configuration.
- Important details
- Cluster naming, automatic node names vs manually “renamed”
- SVMs youll need, protocols they will use
- DNS Server, default domains
2.5 Plan host configuration.
- Protocols needed?
- Multipathing software installed?
- Snapdrive and Host Utilities installed (FC or iSCSI)?
- Verify paths
- HBAnywhere (Emulex) or SANsurfer (Qlogic) for FC installed?
2.6 Create a Snapshot plan.
- What’s the customers RPO (Recovery Point Objective)?
- It is the maximum tolerable period in which data might be lost from an IT service due to a major incident. The RPO gives systems designers a limit to work to. (Wikipedia)
- What’s the customers RTO (Recovery Time Objective)?
- The targeted duration of time and a service level within which a business process must be restored after a disaster (or disruption) in order to avoid unacceptable consequences associated with a break in business continuity. (Wikipedia)
- Expected Space Reservations needed?
- What’s the expected rate of change? Plan for the needed space withing the volumes for this X the number of snapshots
2.7 Plan Ethernet switch configuration.
- Remember- at least 2 different types of switches
- Management
- Flat lights out management switch config. Single vlan. (See section 3 for example)
- Cluster Interconnect
- These have specific configurations based on the number of cluster nodes. (See section 3 for example)
- Data (management may be on these normal switches)
- VLANs needed? Any routing?
- iSCSI
- NFS
- Management / iLO?
- User desktops
- Servers (Web/App/DB tiers)
- MTU?
- Jumbo: 9000 MTU? (needed for FCoE)
- Normal: 1500 MTU?
- QoS?
2.8 Plan zoning configuration.
- Do you have 2 fabrics? 2 FC switches?
- Are these fabrics split or ISL trunked together?
- Multiple ports from each Host, to 2 or more FC switch?
- Verify 4+ paths from each host to the controllers
- Multipathing software:
2.9 Plan iSCSI configuration.
- Gather all IQNs, which iGroups will they belong to?
- Add multiple IP addresses, one for each LIF to the discovery and static bindings
- Enable MCS (Multiple Connections per Session)
- iSCSI boot from SAN needed? (May influence dynamic vs static persistence in iSCSI discovery, and also host LUN IDs used)
- Use VLANs to ISOLATE Traffic!
- CHAP or IPSEC needed?
