Reliability; Availability; Serviceability - IBM 3592 Maintenance Information

Reliability

• Fewer cables because of integral canister plug-in enclosure
• Reliable, brushless motors
• Continuous channel calibration for improved data integrity
• Durable cartridge loader design
• Drive canister is a single FRU
• No pneumatics
• Reliable electronics because of one card versus multiple cards
• Speed matching that requires less backhitching
• No mechanical adjustments on the drive
• Low power requirements
• Proven microcode that is based on predecessor products
• Simplified tape path and tape tracking

Availability

• Drives and power supplies are hot-pluggable, meaning a single drive can be removed or installed while
power is still applied to the product, without affecting the remainder of the product or the customer's
operations
• Product is designed for first-time fix (reduces CE callbacks)
• Product has built-in hardware/microcode error detection and isolation
• Captured error that is recorded in expanded nonvolatile Vital Product Data (VPD)
• Reduced number of FRUs
• Redundant power supplies
• Concurrent microcode update, new/old copy of drive microcode - switch during inactivity
• Service Information Message/Media Information Message (SIM/MIM) clearly indicate impact of failures,
FRUs affected, and the impact of service
• No scheduled downtime that is required for preventive maintenance
• Two CE modes (one allows review of logs and status without customer job interruption)
• FRUs designed for quick and easy replacement
• Product has high level of integration with fewer subassemblies
• Predictive maintenance is possible with SIM/MIM and TapeAlerts
• The product uses a sophisticated algorithm called SARS (Statistical Analysis Reporting System) to track
error statistics to isolate between drive and media problems.
• A warning message is given when the hardware or media is degraded
• The design eliminates the need to take multiple drives offline for service

Serviceability

• Service actions can be initiated from the virtual service panel.
• Wrap tools are provided to test the Fibre Channel external interface
• Automatic dump to flash memory
• Non-disruptive microcode dump
• Drive microcode can be written into the flash memory, when the drive is IMLed, without affecting the
current microcode in DRAM
• No error recreation that is required to isolate the failing FRU
• Error information (SIM/MIM, TapeAlert, Request Sense data) is sent to the host
Introduction 21