Six months after install, a third expansion shelf arrives and the tech lands it on a free port mid-chain without consulting any documentation. Which set of rules did this installation skip?
Select an answer to reveal the explanation.
Short Explanation
The rules you followed for shelf one are the rules for shelf three. Land it in 'any free port' and you get a chain that may work today and fights you every time after — wrong port direction, surprising power sequencing, and a map that now tells lies. Re-read the chain rules before the new shelf goes in.
Full Explanation
Expansion after the initial install re-uses the installation's rules in full. The new shelf joins at the documented position in the chain through the designated in/out ports rather than any free socket, its power cords land on separate A and B feeds and come up in the planned sequence so the head re-enumerates the chain cleanly, and the SAS map is updated so the as-built still describes reality. A free port mid-chain invites an undocumented topology that works by luck and defeats later troubleshooting. A full head shutdown per added shelf is not the skipped rule: downtime windows are scheduled by the site, but even a planned change follows the chain rules of position, ports, power, and documentation, none of which were consulted here. Serial inheritance fails by concept — serial numbers are factory-set, and what the record tracks is chain position and enclosure identity, neither of which anything inherits. Mandatory re-cabling of the whole chain with fresh assemblies is invented: existing interconnects stay in service unless defective or the map changes. Exam caveat: treat an expansion as a rerun of the cabling task, not an improvised addition. Operational check: the chain matches the updated map, the new shelf enumerates through both of its paths, and both of its power feeds sit on separate PDUs.