3Ware 8006-2LP Manual

Contents

Troubleshooting: Problems and Solutions104 3ware Escalade ATA RAID Controller User GuideAEN 02Eh AEN_REPLACE_DRIVE_TOO_SMALLThe ATA RAID controller notifies the user by this AEN when the replacementdrive capacity is smaller than required. The replacement drive must be equalto or greater capacity than the drive it's replacing. For more information referto the Hint on page 32.AEN 02Fh AEN_UNIT_NOT_ZEROEDThe ATA RAID controller notifies the user by this AEN when verificationfails because the array has not been initialized. This is normal for the first ver-ification that occurs on a RAID 1 or RAID 10 array.AEN 021h AEN_UDMA_DOWNGRADEThe ATA RAID controller communicates to the ATA disk drives through theUltra DMA (UDMA) protocol. This protocol ensures data integrity across theATA cable by appending a Cyclical Redundancy Check (CRC) for all ATAdata that is transferred. If the data becomes corrupted between the drive andthe ATA RAID controller (e.g., an intermittent cable connection) the ATARAID controller detects this as a UDMA CRC or cable error. The ATA RAIDcontroller then retries the failed command three times at the current UDMAtransfer rate. If the error persists, it lowers the UDMA transfer rate (e.g., fromUDMA 100 to UDMA 66) and retries another three times. This AEN is sent tothe user when the ATA RAID controller lowers the UDMA transfer rate.Possible causes of UDMA CRC errors are intermittent or bad ATA interfacecables or cable routing problems through electrically noisy environments(e.g., cables are too close to the power supply).AEN 023h AEN_SECTOR_REPAIRThe ATA RAID controller supports a feature called dynamic sector repair toallow the unit to recover from certain drive errors that would normally resultin a degraded array situation. For redundant arrays such as RAID 1, 10 and 5,the ATA RAID controller essentially has two copies of the users data avail-able. If a read command to a sector on a disk drive results in an error, it revertsto the redundant copy in order to satisfy the host’s request. At this point, theATA RAID controller has a good copy of the requested data in its cache mem-ory. It will then use this data to force the failing drive to reallocate the bad sec-tor, which essentially repairs the sector. When a sector repair occurs, the useris notified by this AEN.The fact that a sector repair AEN has been sent to the user is an indication ofthe presence of grown defects on a particular drive. While typical modern diskdrives are designed to allow several hundred grown defects, special attentionshould be paid to any drive in an array that begins to indicate sector repairmessages. This may be an indication of a drive that is beginning to fail. Theuser may wish to replace the drive, especially if the number of sector repairerrors exceeds 3 per month.