Category: Failure Teardown

Part numbers: 00Y5860 / 00RY384 (IBM Storwize V5000 node canister) An afternoon in early August, a data centre called it in, and the description was vague: “Nothing is down, but it’s very slow.” The array was online, the volumes were all there, hosts could read and write, and there was no outage alarm at all.…

Part number: FAS8200 (NetApp controller) This is where a NetApp controller replacement most often stalls, and the symptom when it does is alarming: the part is in, the machine came up, the indicators are normal, but the new controller cannot see its half of the disks. The spare is not faulty. A step was missed.…