I have an older HPE Nimble CS3000 ES2-H21T test array with two controllers. Controller A developed a boot-loop problem, and the array did not automatically fail over to Controller B. Controller B still responds to ping, but I cannot connect to it over SSH. I have replaced Controller A with another controller, but the system is still not operational. This array is no longer covered by Nimble support. Since it is only a test setup, I am willing to rebuild it from scratch. What is the correct and safe process for recovering the controllers, resetting or formatting the array if necessary, and configuring it again?
2 Answers
Treat this as proprietary enterprise storage rather than a server that can simply be wiped and reinstalled. Before attempting any reset, verify that anything important is backed up. If the array has an active support contract, HPE is the safest—and likely only fully supported—source for controller replacement, firmware matching, and recovery instructions. Without support, replacing the array with a currently supported lab platform may be more realistic than trying to reconstruct the system from scratch.
The CS3000 is an end-of-service platform, so the normal recovery path may no longer be available. Nimble arrays generally require the supported management interfaces and matching firmware before controllers can operate together; firmware usually cannot just be loaded manually. Diagnostic and recovery access may also be restricted to HPE or Nimble support. If this is only a lab system, consider whether controller failover is actually necessary, since rebuilding the array may not be practical without the appropriate tools or support entitlement.
Even though the hardware is fundamentally a computer, these arrays depend on vendor-specific firmware, controller pairing, and management procedures. That makes unsupported experimentation risky and much less straightforward than reinstalling a normal server.