|
|
1U Raid5 --- Laptop Raid5
|
- RAID -- Redundant Array of Inexpensive Disks
- RAID is NOT a backup system ...
Please backup all your important data properly
- RAID1 allows you to boot from the other mirrored disk for some cases of hard disk failures
- RAID5 protects your data against one hard disk failures
|
1U Raid5[tm] Servers
|
- 8 drives at 200GB per disks ===>> 1.6TeraByte 1U Raid5[tm] Servers
|
Laptop Raid5[tm] Servers
|
- Up to 12 laptop drives at 80GB per disks ===>> 960GigaByte Laptop Raid5[tm] Servers ( 1U Chassis )
|
SW Raid uHow2
|
- Apply the raid-2.x.y patches to linux-2.0.x and linux-2.2.x kernels
- raid0 works out of the box
- 2.0.x and 2.2.x kernels
- raid1, raid4, raid5 requires patches
- 2.4.x kernels
- raid5 works out of the box
- Download and install the Older raidtools if you need it ( local copy )
- ckraid, mdcreate, mdadd, mdrun, mdstop
- Download and install the current raidtools ( local copy )
- mkraid, raidstart, raidstop
- Download and install the latest IDE patches if needed ( local copy )
- Create /etc/raidtab file
- RaidStart, mke2fs, mount /Raid5 Commands
- Backup important Raid5 data
|
| Testing the Raid SubSystem |
- Simulate a hard disk failure
( unplug it or take it out with raidhotremove )
- Create a known to be good 4GB test file
- Create lots of 4KB, 400KB, 4MB, 400MB, 4GB, 40GB sized files to fill the raid subsystem
- the copied file should be the same as the original file
- You should NOT have to touch anything to boot the raid box
- Power Off and Pull the disk out ( /dev/hda )
- a properly configured raid system will boot without any intervention
- copy and compare the Good.4GB.file to the degraded mode raid subsystem
- Power back up normally and let /dev/hda resync itself ( catchup )
- it should finish copying the test 4GB file to /dev/hda
- Power Off and Pull the OTHER disk out ( /dev/hdc )
- a properly configured raid system will boot without any intervention
- copy and compare the Good.4GB.file to the degraded mode raid subsystem
- Power back up normally and let /dev/hdc resync itself ( catchup )
- it should finish copying the test 4GB file to /dev/hdc
- Power Off and REPLACE /dev/hda with a brand new disk
- a properly configured raid system will boot without any intervention
- Restore the RAID substystem to "normal operations"
- System finished resyncing ( cat /proc/mdstat )
- Repeat the process for each disk in the raid array .. till you're tired/confident :-)
|
|
|