This chapter lists error messages that you might see on a system with one or more RAIDs installed and RAID LED error conditions that can be seen after opening the RAID unit. The causes and possible solutions of these errors are explained. This chapter contains these sections:
Section 5.1, “raid Error and Warning Messages,” lists messages from the raid(1M) command. These messages appear as standard output and optionally on the console and in /var/adm/SYSLOG.
Section 5.2, “Messages from the RAID Device Driver,” lists messages from the RAID device driver (see usraid(7M)). These messages always appear on the console and in /var/adm/SYSLOG.
Section 5.3, “LED Error Conditions,” describes error conditions that are indicated by LEDs in the RAID unit.
Error messages from the raid command are always displayed in your shell window. When the raid –L option is given, messages are also written to the console and /var/adm/SYSLOG.
drive n is missing, formatting failed drive n is not the same size as the other drives: m sectors drive n is not a supported drive type: vendor: string1, product: string2, revision: string3 |
All disk drives must be present, working correctly, and of the same make/model.
/dev/rdsk/radcontrollerdunitvh: RAIDlevel, sectors: sectors, stripe size: depth, drive n is down /dev/rdsk/radcontrollerdunitvh: RAIDlevel, sectors: sectors, stripe size: depth, not initialized /dev/rdsk/radcontrollerdunitvh: RAIDlevel, sectors: sectors, stripe size: depth, must reformat |
The RAID is not formatted properly. Try steps 7 through 9 of Chapter 2 again. If you get the error message again, there is a hardware problem. Contact your local service provider.
cannot check integrity, drive n is down |
All disk drives must be up and functioning. There is no point in checking the parity integrity of the RAID when a disk drive is down.
controller has been replaced, it doesn't match disk config info |
If the controller has been replaced, then any indication of down disk drives is not valid. The –m option wasn't specified, so no action is taken.
controller has been replaced, reprogramming to match disks |
If the controller has been replaced, then any indication of down disk drives is not valid. The –m option was specified, so the controller configuration will be forced to match that of the disk drives in the RAID.
disk n has been replaced, but it wasn't marked down |
This can happen if a disk drive is replaced while the system was down.
disk n is being forced down, the RAID needs to be rebuilt |
The –m option was specified, so the disk drive is being marked down. It must be rebuilt using the –r option.
drive n is down |
The disk drive has failed or been marked down. It it has failed, it should be replaced and rebuilt. If it was simply marked down, it can just be rebuilt.
can't download firmware to an uninitialized RAID, re-format (-3|-5) the RAID, then try again |
The RAID must be formatted and in good working order before new firmware can be downloaded.
cannot force a drive down on this RAID |
There is some problem with the RAID. A disk drive can't be forced down at this time.
the controller has been replaced. use the `-c' and `-m' arguments to reprogram the controller to match the config info on the drives |
If the controller has been replaced, then any indication of down disk drives is not valid. Using the –c and –m options forces the controller configuration to match that of the disk drives in the RAID. After that is done, disk drives can be marked down using the –d option, if desired.
disk n has been replaced, but it wasn't marked down |
This can happen if a disk drive is replaced while the system was down. The –c and –m options are used at system boot time to detect this situation and mark any such disk drives down. They must be rebuilt using the –r option.
aborting: this would down a second disk in the array |
Every attempt is made to avoid marking a second disk drive down, but the –f option will forcibly mark any disk drive down. If a second disk drive is marked down, the RAID must be reformatted before it can be used for anything. All data will be lost.
cannot rebuild this RAID all disks are OK in this RAID |
There is either a problem with the RAID and it cannot be rebuilt at this time, or there are no disk drives marked down.
drive n is missing, rebuild failed drive n is not the same size as the other drives: m sectors drive n is not a supported drive type: vendor: string1, product: string2, revision: string3 |
All disk drives must be present, working correctly, and of the same make/model.
more than one disk is marked down |
More than one disk drive has been marked down. This can only happen when an administrator forces down more than one disk drive. The only recovery is to reformat the RAID.
drive n is missing transfer of config info to/from disk n failed |
There is configuration information stored in reserved sectors on each disk drive in a RAID. That information could not be read from the indicated disk drive. Either the disk drive has totally failed or the disk drive is not plugged in. This message will probably be accompanied by messages from the kernel.
drive n's config info is unrecognizable drive n's config info contains illegal values disk n config info doesn't match NVRAM info |
These are all indications that something is wrong with the disk drive in the given slot number.
drive n should be in slot m, raid ignored |
The configuration information from the reserved sectors of each disk drive shows that the disk drives in the RAID have been moved around.
controller and drives disagree on which drive failed, controller says n failed, while drives say m failed |
The configuration information from the reserved sectors of each disk drive is inconsistent. It is unclear which situation is true. Depending upon which situation is true, all the data in the RAID could be lost. The controller's NVRAM claims that one disk drive failed, and the configuration information in the reserved sectors of that disk drive match all the other disk drives, but there is a second disk drive whose configuration information does not match. Probably a double failure has happened, and the RAID must be reformatted and reloaded from a backup tape.
raid corrupted, must reformat |
The internal consistency of the RAID parity information has been compromised with no possibility of recovery. The RAID must be reformatted (wiping out all data on the RAID), and the contents reloaded from a backup tape.
raid not initialized |
Formatting was not successfully completed. The RAID is not yet usable; it must be reformatted.
All messages from the RAID device driver begin with:
radcontrollerdunitspartition |
RAID device driver messages are written to the console and /var/adm/SYSLOG.
disk n has failed, running in degraded mode |
The disk drive in slot n has failed; it should be replaced as soon as possible.
disk n is predicting failure, replace as soon as possible |
The disk drive in slot n has returned a PFA warning. It will probably fail within 24 hours. It should be replaced as soon as possible. You should probably mark it down before replacing it, but that is optional.
not ready, initializing command required |
The controller has not been configured, the configuration has been zeroed (with the raid –z option), or the controller configuration has been corrupted somehow. If the controller had previously been configured, try resetting the controller by pushing the controller Reset button (shown in Figure 1-5), otherwise a format operation is required.
unrecovered hardware error, possibly a drive is not responding |
A disk drive is not responding, but the controller cannot tell if the disk drive is dead, simply not plugged in, or there is a fault in the bus connecting the controller to the disk drive. Check all the disk drives in the RAID unit.
fixing parity integrity on stripe at LBA n |
An integrity check operation has detected that parity doesn't match the corresponding data for the stripe starting at the given block address. The parity is being updated to match the data automatically. If large numbers of these appear, or they recur often, contact your local service provider.
recovered error |
A hard error was detected on one of the disk drives in the RAID, but the controller has reconstructed the data and taken recovery actions.
volume header not valid |
The volume header on the RAID is not valid. fx(1M) should be used to write a new one.
retrying request retries exhausted |
A command has had an error and is being retried. After a certain number of retries, IRIX gives up and returns an error to the user.
fatal error on maintenance command maintenance command LBA not changing, controller may need to be RESET unable to abort maintenance command, controller may need to be RESET |
A maintenance command (initialize disk drive, integrity check, or rebuild disk drive) has had an unexpected error. Retry the operation.
forcing a second drive down |
The raid command was used to forcibly mark down a disk drive in a RAID where there was already one disk drive down. Data may have been lost.
FLASH EPROM programming error FLASH EPROM sector protection error at page n microcode download failed new microcode has invalid checksum, ignored new microcode has invalid signature, ignored |
There was an error downloading firmware to the controller. Retry the operation.
auxiliary input 1 active auxiliary input 2 active |
There are auxiliary inputs on the controller for monitoring internal conditions. Contact your local service provider.
controller had internal parity error internal target failure unrecovered hardware error unrecovered hardware error, 2nd fault while recovering unrecovered hardware error, cannot find task ptr unrecovered hardware error, recovery steps skipped unrecovered hardware error, unknown cause unrecovered hardware error, which stream is unknown unspecified hardware error |
The controller has reported an internal hardware fault. Use fx to run diagnostics, then contact your local service provider.
diagnostics failed on SCSI channel n: buffer test diagnostics failed on SCSI chip channel n at ID m diagnostics failed |
Diagnostics were run on the controller and a hard fault was detected. Contact your local service provider.
corrupted definition page corrupted partition page corrupted units down info |
The RAID configuration information (the RAID level, stripe depth, and failed disk drive information) returned by the controller is not in a recognizable form. Try resetting the RAID (Section 4.3, “Restarting a Hung RAID”). If that doesn't work, try reformatting (see Chapter 2, “Formatting a RAID”). If that doesn't work, use the –z option to zero out the controller configuration (see Section 4.4, “Resetting a RAID to Factory Defaults”). Contact your local service provider.
RAID controller LEDs display these error conditions:
LED 2 is not flashing.
When LED 2 is not flashing (the flashing cycle can be very slow, up to 10 seconds), the RAID has failed, possibly due to an unrecoverable hardware error. Try both procedures in Section 4.3, “Restarting a Hung RAID,” to see if the RAID will start operating again. A service call by your support provider may be required to diagnose and repair the problem.
Any of LEDs 4 to 8 are lit.
Each lit LED indicates that a disk drive has failed. See Table 1-2 and Figure 1-5 to figure out which one. Chapter 4, “Recovering after a Disk Warning or Failure,” explains how to fix the problem.
When the amber LED in a RAID disk drive is on (see Figure 1-6 for the location of the amber LED), that disk drive is down. It could be down because it has failed or because it has been manually marked down with the raid –d command. See Chapter 4 for recovery procedures.