At 17:26 15/08/2002 +0800, you wrote: >Hi all folks, > >I am going to experiment software RAID 5. Any folks on the list have >experience and what are your comment ? > >Proposed setup :- >Hardware available >2 ATA133 hard discs >1 ATA100 hard disc > >1 ATA 133 hard disc connected to IDE (primary) slot of the motherboard as >MASTER as parity 1 & 2 (block 3 & 5) > >1 ATA 133 hard disc connected to IDE (primary) slot of the motherboard as >SLAVE as parity 3 & 4 (block 1 & 6) > >1 ATA '100' hard disc connected to secondary IDE slot of the motherboard >as MASTER as parity 5 & 6 (block 2 & 4) > >Will above connection be OK? If hard discs of different spec are used, >does it affect the RAID performance? > >Your advice would be appreciated. Thanks in advance
A few points:- 1) The array will be bottlenecked by the slowest drive. 2) The total size of the arrays is dependant on the smallest drive. 2) Running two drives on the same interface will cause severe performance degradation. 3) The array with /boot on it must be raid0 or 1 - raid5 does not work for bootloaders. 4) This should be fine if you just want to experiment - there's a lot to work out with software raid when it comes to configuration files and testing for and replacing failed drives without losing data. I run software raid-5 pretty much everywhere so throw me a line if you feel the need. There's a software raid monitor tool at "ftp://ftp.nexnix.co.uk/pub/linux/scripts/mdmon"; if you should need it. hih nick@nexnix >Stephen Liu > > > > >-- >redhat-list mailing list >unsubscribe mailto:[EMAIL PROTECTED]?subject=unsubscribe >https://listman.redhat.com/mailman/listinfo/redhat-list -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED]?subject=unsubscribe https://listman.redhat.com/mailman/listinfo/redhat-list
