Date: Wed, 27 Mar 2002 22:40:35 -0800 From: "frank amo" <lxfrank90@hotmail.com> To: questions@FreeBSD.org Subject: Mylex dac960pl raid "system hangs during installation" FreeBSD 4.4 Message-ID: <F73RBaYFkuutSgcOrBL00001d8f@hotmail.com>
next in thread | raw e-mail | index | archive | help
Im currently in the hardest linux/unix crisis in my linux life.
I cannot install any linux/or/freebsd distros on my server intended
computers. Im going to break down these paragraphs into outlines starting
with the hardware description, paragraph, operating systems, paragraph, some
questionss, and collective ideas and tips from other linux support people.
Server Hardware:
CPU = dual intel pII 333mnz
BUS = 66mhz
RAM = 1gb sdram pc100
IDE = floppy/cdrom/(ofcourse)
RAID= MYLEX DAC960pl scsi raid controller
bios = 4.xx-x
scsiRAM= 16mb (detected)
write policy = write-back(write-thru->default)
firmware=
originally 3.50-xx(problematic)
upgraded to 3.52-x-xx(lessProblems)
drives initialized
target 1 and 2 detected on channel 0
packs created and drives initialized with firmware upgrade
1 system drive installed (both ibm drives) as 700001mb
SCSI cable terminated with an lvd terminator.
HARDDRIVES= 2 ibm ultrastar scsi harddrives 10000rpm 36.x GB each
NETWORK= 3com 3c905 ethernet
DEVICES= AMI MegaRAC remote access companion adapter with pcmcia
Operating Systems:
Linux= RedHat 7.1, SuSE 7.1, SlackWare 7.1 & 8.1
Unix = FreeBSD 4.4, and 4.5
DOS = 6.22
Original Problem:
With old firmware
+++
+++RedHat 7.1
The installation would not go past the detection of the mylex raid
controller. It would fail. "No harddrives installed"
"kernel panics"
"kernel traps"
"unable to handle virtual address ax00000xx ("x" is any number)"
"signal 11 (I recall seen every other error) "
Ive even seen red screens of death, with the above messages.
++++
++++SuSE 7.1
The installation would not go past the detection of the mylex raid
controller. It would fail. "No harddrives isntalled"
System would crash, with similar errors as RedHat but not as dramatic.
+++++
+++++SlackWare: 7.1 & 8.1 same problem.
I did not get too far because I always thought that the scsi drives were
refered to as "/dev/sda" not "/dev/c0d0xxx(whatever)" which I found in
RedHat 7.1 described in the "New Problem Section" using "fdisk" and
"disk-druid".
SlackWare did report the mylex controller in the /proc/pci file. So the
kernel does detect that the harddware is there.
+++++
+++++FreeBSD 4.4 & 4.5
During bootup I would see a message in regards to the Mylex controller that:
mxld0 "mylex" this firmware is not recommended.
The system would recognize both of my harddrives, allow me to "create
slices" (partitions) which I did auto, allow me to lable my partitions or
give them mount points, which I did automatically.
The "newfs -f x 1024" begins formating my harddrives as I check in the other
tty and the installation begins.
Here to after a certain percentage, I encounter almost identical error
messages to the redhat 7.1 error, the system set itself to reboot in 15
seconds in order to clear itself from that error.
(one wierd thing that happened on the highest percentage of installation
without that error the system hung during installation as usual, I rebooted
to the SCSI(harddrives) and the system then booted to a weird "ok" prompt.
heres what it looks like:
----------------------------------------------------
ok ? (lists about 10 - 13 commands)
ok ls (lists the basic directories)
ok cd (does not work here)
ok load (I tried "load /modules/module.o) (useless)
----------------------------------------------------
Ill go to DOS:
DOS installed with no problems. (but its only dos and I dont want to install
linux on a fat16 or fat32 partion.
New Problem:
After Mylex Firmware and Bios upgrade:
Now only testing with RedHat 7.1 and FreeBSD 4.4
RedHat 7.1
I have not recieved any of the above errors yet.
I used the default graphical installaation, text installation, and even
expert installation.(which is where I discovered that the scsi harddrives
have a name that I havent memorized /dev/c0f0c1(whatever) not /dev/sda.
The problem now is that after defining my partitions with either fdisk or
disk druid, I select my mount points set my <swap> to 2047 since I have 1
gigabyte of memory.
When I begin the installation here is what happens.
1. Partions and filesystems are formated.
2. Install image gets transferred to the harddisk.
3. Very few packages begin to install.
4. System freezes for up to 35 minutes or until I get sick of waiting, and
yes I did a mickey mouse test by either checking my numlock or mouse
scrolling to check if it froze and after second numlock keystroke I find
that the light does not turn off or on depending on where it lef off.
FreeBSD 4.4
The error regarding the mylex firmware is gone. Everythings appears without
error.
The installation went further than ever.
The system hung after formatting the /usr/local file systems.
I got sick of it and started all over again.
Here are my assumptions.
1.) Bad mylex drivers? No. Because the systems recognize them and they do
not reject them. By the way Could someone please tell me how could any one
load a kernel module for the dac960pl (ifneeded) when I dont even have linux
installed on the harddrive yet.
2.) Kernels 2.2.16, 2.2.18, 2.4.x do not really support the mylex dac960raid
controller.
3.) Incorrect scsi raid termination? Then how would the mylex configuration
utility recognize the harddrives as target 1 and target 2
on channel 0 as one 70001 system driver and individually at 3500x each
drive? No to rule out that DOS was able to install successfully.
4.) Bad memory? You can be right. BIOS counts it up, but on other linux
systems that I have I corrected similar problems by removing most of my
dimms and corrected two different computers by removing the apparent bad
memory? What confuses me is that I have not seen any of those errors since I
upgraded the firmware/bios on the mylex controller.
5.) CPU handling errors. I selected the "individual packages" on RedHat 7.1
installation and selected the "smp-kernel" option.
6.) I saw something in the RedHat Manual regarding CPU cache. My Processors
have 512k cache in each. When I disabled the "internal-cache" in the bios
this morning, RedHat got stuck on the /sbin/loader before installing.
7.) ?
8.) ?
9.) oops!!! mybrain=0
10.) frank panic!!!
I have two identical computers as described above. The only difference
between the two is the color of the cpu heatsinks. I have never encountered
an error with any other Linux installations this grave.
Its almost impossible that two computers can have the same identical problem
for the same identical cause.
What do I do?
How do I doit?
Please help.
Thanks
Frank
_________________________________________________________________
Join the world’s largest e-mail service with MSN Hotmail.
http://www.hotmail.com
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?F73RBaYFkuutSgcOrBL00001d8f>