[ale] RAID reconstruction

David Corbin dcorbin at machturtle.com
Sun Nov 24 18:25:09 EST 2002


speed=6155K/sec is what the /proc/mdstat says.  Seeing as that is more than any number set in my system, I don't really know how to interpret it (too fast, or way too slow).  This is an two-drive IDE system, for reference....I kind of figured that since it is IDE, I'd see a lot more CPU utilization if it was I/O bound (It's about 5% CPU on a Pentium 200).




Robert L. Harris wrote:

>Then you're likely IO bound already.  My scsi systmes seem to max out
>around 180-220 what is your at when it's set to 500000 and 100000?
>
>Thus spake David Corbin (dcorbin at machturtle.com):
>
>  
>
>>Date: Sat, 23 Nov 2002 19:27:01 -0500
>>From: David Corbin <dcorbin at machturtle.com>
>>To: "Robert L. Harris" <Robert.L.Harris at rdlg.net>
>>CC: ale at ale.org
>>Subject: Re: [ale] RAID reconstruction
>>
>>That doesn't seem to have much affect on how quickly it reconstructs 
>>(though you'd certainly think it would).
>>
>>
>>Robert L. Harris wrote:
>>
>>    
>>
>>>cd to /proc/sys/dev/raid and check out the speed* (don't have access to
>>>a system running them at home)
>>>
>>>you can echo a # to them, I usualy set them to 500000 and 100000,
>>>that'll set a pretty high min and over the physical limit max.  It'll
>>>resync very quickly but can really hammer your cpu, you'll want to
>>>reduce it when the sync is done.
>>>
>>>
>>>Thus spake David Corbin (dcorbin at machturtle.com):
>>>
>>>
>>>
>>>      
>>>
>>>>Date: Sat, 23 Nov 2002 14:20:35 -0500
>>>>From: David Corbin <dcorbin at machturtle.com>
>>>>To: ale at ale.org
>>>>Subject: [ale] RAID reconstruction
>>>>
>>>>When a RAID-1 "drive" reconstructs itself after it is documented as 
>>>>doing no more than 100K/sec/disk, so as not to impact other process to 
>>>>heavily.  Is there anyway to bump this level up?
>>>>
>>>>Thanks
>>>>David
>>>>
>>>>
>>>>---
>>>>This message has been sent through the ALE general discussion list.
>>>>See http://www.ale.org/mailing-lists.shtml for more info. Problems should 
>>>>be sent to listmaster at ale dot org.
>>>>  
>>>>
>>>>        
>>>>
>>>
>>>:wq!
>>>---------------------------------------------------------------------------
>>>Robert L. Harris                     | PGP Key ID: FC96D405
>>>                             
>>>DISCLAIMER:
>>>    These are MY OPINIONS ALONE.  I speak for no-one else.
>>>FYI:
>>>perl -e 'print $i=pack(c5,(41*2),sqrt(7056),(unpack(c,H)-2),oct(115),10);'
>>>
>>>
>>>
>>>      
>>>
>
>
>
>:wq!
>---------------------------------------------------------------------------
>Robert L. Harris                     | PGP Key ID: FC96D405
>                               
>DISCLAIMER:
>      These are MY OPINIONS ALONE.  I speak for no-one else.
>FYI:
> perl -e 'print $i=pack(c5,(41*2),sqrt(7056),(unpack(c,H)-2),oct(115),10);'
>
>
>---
>This message has been sent through the ALE general discussion list.
>See http://www.ale.org/mailing-lists.shtml for more info. Problems should be 
>sent to listmaster at ale dot org.
>
>  
>

_______________________________________________
Ale mailing list
Ale at ale.org
http://www.ale.org/mailman/listinfo/ale






More information about the Ale mailing list