[Samba] AutoCad and ArcView

[James McLaughlin]

Jean-Sebastien Pilon wrote:
> 1. Are the files constantly accessed by the applications ( like an excel
> file opened on a share ) ?
Yup
> 
> 2. What RAID settings does your controller support other than 5? 
>    I know IBM servers use 5E, 5EE and 1E which are a little more
> interesting than plain 5
Vanilla Controller 1/5

>    A raid 10 could be interesting too for file serving where writes are
> as common as reads.
> 
After doing some research I found you probably hit the nail right on the
head for our RAID config.  Raid 5 is great for smaller files.  Where as
RAID 10 is designed for large file transfers.  
We recently acquired a new adaptec controller The (ASR2130SLP) which
does support RAID 10.  
######################################
>On Wed, 2006-11-15 at 13:05 -0500, Aaron Kincer wrote:
>> Can you be more specific than "burned through"?

Hard Drive Failed. error 32. -->Multi-bit ECC / Both Drives 

>> What brand of hardware are you running?

5 Fujitsu 320GB Drives
Intel Server Board SE7501CW2
Vanilla Intel Controller Card -- I will look up the #'s later.  Sorry

>> I am unconvinced that your OS and/or Samba are the cause of your 
hardware issues. 

Never would I think that -- just wondering if anyone else had been using
a SAMBA server in a like environment and had some information on how
they had set up a smooth running server.
These things are like cars.  Lets get under the hood and do some
tweaking -- knowwhatImean?  Nothing wrong with adding a little extra pep
by tweaking conf files, hosts lists, remove bottlenecks, etc...etc...

>> I've seen data intensive servers running RAID 5 run for 
many years with no problems.

Our backup server for this system is 1/2 the machine and ran fine for 2
weeks while we waited on hardware.  Quite a bit slower and lots of
complaints about access speeds etc...but nary an issue.

> I would be more inclined to believe you've 
got a faulty backplane that is providing irregular power that might be 
zapping your drives than software related.

Hmmm....are there any packages out there that can be installed to
monitor this? 
Anyone know if Nagios monitors items of that detail -- or have the
ability?  I will havet to look into that. 

> Out of curiosity--what is the same slot in your array that suffered
the 
failed drive?
Magic Number is 4

JM