[CentOS] expert guideance needed on arrays with CentOS 4 installed on dual processor compaq server

Robert Hanson

roberth at abbacomm.net
Sun Aug 14 15:26:23 UTC 2005


greetings,

as most know i have some older dual processor Compaq playtoys that i am
deploying CentOS 4 on and doing some real world'ish simulation of failure
and regular server environment testing.

i have searched www on some things to no avail so off to the experts for
some guideance please?

case of:

a dual Pentium 3 processor 1850R with a smart array 221 controller and
quantity (4) 18 Gig SCSI hard drives in default specs RAID5 array created by
the controller and Compaq Smartstart CD and i set the RAID rebuild to
"highest priority"...

Then, i do a custom install without Gnome or KDE of course.

linux text lowres nousb nousbstorage skipddc

install CentOS 4 using autopartition with a 

/boot 	1024MB i.e. 1Gig
/		approx 50,000MB i.e. 50Gig

everything installed perfect and i 

rpm --import /usr/share/doc/centos-release-4/RPM-GPG-KEY

imported the key and do a full

yum update

to latest greatest CentOS 4 packages....

Now, i am trying to get some perspective and pointers so i can learn more
how to better understand what i really have here. i know i have a lot to
learn in regards to managing RAID5 on CentOS 4

Should I have autopartitioned?
Is LVM the best way to manage RAID5 on CentOS?
What other tools are there on a default CentOS 4 install to manage my array?
What tools should i install with yum install to learn more and help me with
this situation?

If i shutdown the machine, and yank out any drive, and replace it with an
identical drive and fire the unit back up, it asks me if i want to rebuild
the data.

Are these Compaq arrays "hot swap" like i think they are?

How "long" does that usually take for it to rebuild by the controller?
Hours? Days? Months? ;->

Remember, this is a (4) 18gig drive system.

Should I allow the machine to be in production, online, up and running while
it is rebuilding the 4th disk?

Yes, I get the error from the kernel...

non fatal error on ida/c0d0

....while the machine appears to be rebuilding the 4th disk in the array...

What other CentOS software tools should I be aware of to help me here?

I really really appreciate your help and pointers and will go check out all
of the documentation that you folks will point me to.

Please remember, these boxes are not "in production" right now, so this is
not an emergency, yet i am trying to simulate a production environment for
obvious reaasons.

Thank you in advance and kind regards,

- rh

--
Robert Hanson
Abba Communications
http://www.abbacomm.net
 





More information about the CentOS mailing list