|
|
Sponsored Content
Special Forums
Hardware
Filesystems, Disks and Memory
Linux Storage system: looking for advices
Post 302385670 by Loic Domaigne on Saturday 9th of January 2010 03:07:36 AM
01-09-2010
Quote:
Originally Posted by pludi
Which is, usually, something you do not want to happen. Personally, I'd rather have a system that's a bit slower if a disk goes bad, but keeps on working, than losing files and not being able to use the system until I've got a replacement disk. But that might just be me.
Quote:
Originally Posted by Loic Domaigne
However, I failed to restore the file system if /dev/vda6 gets damaged. I used an alternate superblock for fsck (one located on vdb2 or vdb3), but no avail.[...]
Quote:
Originally Posted by pludi
That might be because of the striping done by LVM. The volume manager doesn't fill up the first disk, then the second, and so on, but first fills the first stripe (by default 4 MB) on the first disk, the first stripe on the second disk, and so on. With that scheme it might very well be that all superblocks end up on the same disk.
Thanks you everyone, and especially pludi
. I gathered enough information to make a sound proposal for the storage system!|
|
LEARN ABOUT PHP
addbadsec
addbadsec(1M) System Administration Commands addbadsec(1M) NAME
addbadsec - map out defective disk blocks SYNOPSIS
addbadsec [-p] [ -a blkno [blkno...]] [-f filename] raw_device DESCRIPTION
addbadsec is used by the system administrator to map out bad disk blocks. Normally, these blocks are identified during surface analysis, but occasionally the disk subsystem reports unrecoverable data errors indicating a bad block. A block number reported in this way can be fed directly into addbadsec, and the block will be remapped. addbadsec will first attempt hardware remapping. This is supported on SCSI drives and takes place at the disk hardware level. If the target is an IDE drive, then software remapping is used. In order for software remapping to succeed, the partition must contain an alternate slice and there must be room in this slice to perform the mapping. It should be understood that bad blocks lead to data loss. Remapping a defective block does not repair a damaged file. If a bad block occurs to a disk-resident file system structure such as a superblock, the entire slice might have to be recovered from a backup. OPTIONS
The following options are supported: -a Adds the specified blocks to the hardware or software map. If more than one block number is specified, the entire list should be quoted and block numbers should be separated by white space. -f Adds the specified blocks to the hardware or software map. The bad blocks are listed, one per line, in the specified file. -p Causes addbadsec to print the current software map. The output shows the defective block and the assigned alternate. This option cannot be used to print the hardware map. OPERANDS
The following operand is supported: raw_device The address of the disk drive (see FILES). FILES
The raw device should be /dev/rdsk/c?[t?]d?p0. See disks(1M) for an explanation of SCSI and IDE device naming conventions. ATTRIBUTES
See attributes(5) for descriptions of the following attributes: +-----------------------------+-----------------------------+ | ATTRIBUTE TYPE | ATTRIBUTE VALUE | +-----------------------------+-----------------------------+ |Architecture |x86 | +-----------------------------+-----------------------------+ |Availability |SUNWcsu | +-----------------------------+-----------------------------+ SEE ALSO
disks(1M), diskscan(1M), fdisk(1M), fmthard(1M), format(1M), attributes(5) NOTES
The format(1M) utility is available to format, label, analyze, and repair SCSI disks. This utility is included with the addbadsec, diskscan(1M), fdisk(1M), and fmthard(1M) commands available for x86. To format an IDE disk, use the DOS "format" utility; however, to label, analyze, or repair IDE disks on x86 systems, use the Solaris format(1M) utility. SunOS 5.10 24 Feb 1998 addbadsec(1M)