2.6.2.1 setup wont load?

I had a similar problem with my install. It hung on something like loading ata...pix cant remember. I did a lot of research amd found a forum that said to change the SATA port of the hard drive to one that was 4,5, or 6 on the motherboard. It worked for me.

Jorge

Your lucky you did not do a yum update on a productions system, I just spent the night reloading everything to the previous stable version becuse the yum update left the system unable to boot, so far I've had no reply to my posts.. I also tried to reboot and do a frsh install with the new ISO and same blue screen.

The same problem on HP ML110 server

My checksum is correct and this image are working on virtual pc

bump

Anyone have any other suggestions or ideas?

What box do you have or if custome what type of motherboard?

Just wondering if its a chipset issue we have or something else.

Sucks that none seem to know the issue or want to help. :( We can see that an issue does indeed exists.

Well the 2.6.1.x all worked fine except for network drivers which I installed myself.

So why would a newer kernal remove support? This box is less than a year old so its not old hardware. Now I do have an IDE Raid Fastrack2000 card in it doing mirror but that has always worked and I would think it would still boot and if it would work would just say no hard drives. But it see's the drivers during the boot because I see it probe when the screen is black. Its when it goes to the blue screen, you pick the keyboard country and than it goes to another one, probes the video than that goes away and boom, thats it. Stays blue.

I am not using anything as Sata... Bios is latest one, I have tried to disable everything in the bios, and it makes no diffrence, I have loaded bios defaults, no diffrence, have played with otehr settings no diffrence.

I tried to load setup with noprobe, get same results.

So I just dont get it. This chipset is supported, the video is support, and they claim now the network is native support but yet now it seems to not work at all.

I saw these exact same symptoms during an install on an MSI Wind PC, which has one of the Intel Atom chips in it. Went to the previous ISO image and it installed normally.

In my opinion there's definately something different going on with the most recent ISO, there's been numerous complaints around here about it not installing properly.

John

Yes, there IS something different. NEW KERNEL!!!

Please understand, a new linux kernel DOES NOT always mean exact same support as the previous kernel builds. It generally means broader range of support for products and updated support for legacy products. However you will need to have your legacy products brought up to snuff to work with the new kernel that is included in the tribox ce ISO.

Thats what I am saying though, nothing legacy in the system. Its all brand new as of 2008 except for the raid card. Which DOES work on Cent 5.2 as I have another box being a firewall that has the exact same card with same firmware on it. Also this TB disk DOES detect the hard drives and gets past that point.

As for testing with CentOS stand alone, my underestanding is that its not the same kernal as the TB anymore so wouldn't that not be a valid test as its not the same?

I've had the last 2 trixbox releases fail on some hardware that has worked in the past. Some dell 1650's that have been in operation since A@H 1.5 have been giving us some problems and some Dell 2850 machines. The problem we are seeing that these machines with the Digium cards in them are causing problems for us. The installation works, although we see errors during the install, and specifically we see asterisk attempting to start and failing. We think that its some sort of dependency/race condition that isn't set up properly during the setup of the installation.
We generally will get to a command prompt after the install, and after we reboot, we have to re-run the /var/trixbox_load/firstboot/91_pbxconfig_setup.sh to get everything to finish up. (we usually see that FreePBX isn't available to us at all until that script is re-run)

Try hitting "alt-F3" and "alt-f4" once it appears to hang, this will take you to different console screens that will show you some of the events happening behind the scenes. There might be something of value in that output to help you track down your issue.

Might want to take a look here: http://trixbox.org/devblog/trixbox-ce-262-released#comments
There's apparently some issues with certain Realtek nic drivers.

Have you tried disabling things such as LAN and Audio controllers in your BIOS before installing?

John

I have tried to disable everything and tried diffrent combos of things disabled. Get same results.

Well I tried the unplug the nic part just now, no go still...

Here is what it stops at when I look at the other screens:

alt-F3 has the last line as:

ISCSIADM is /usr/sbin/iscsiadm

alt-F4 screen has last line:

device-mapper: multipath emc: version 0.0.3 loaded

So any thoughts?

I had a similar problem. Apparently, CentOS will format your HDDs after this step.

My problems were caused by an Intel ICH9R controller onboard.

The solution was to set the following in BIOS:
Change SATA mode to SATA
Switch *off* SATA AHCI

It installed without any further problems.

how do u install network driver off a dvd rom i put them on it??

2.6.1 worked fine, and got blue screen on 2.6.2 ...
so, i conclude, the easyest solution is to change hardware? )) linux, linux...

i've been going nuts tonight trying to install 2.6.2.2 and the same issue described above is happening to me

i know the media is good, as i have done many VM installs and checksumd it

it has gotten to the point where i have disabled everything in the bios that i can, except for primary ide (cd) and secondary ide (target hd), i've even disabled lan, audio, sata, ide raid (fastrak)

my mobo is a gigabyte GA-8INXP, not too terribly old ;)

i can make it to the step where I choose my language (or keyboard) whatever, its US ... and then blank blue screen ... alt f3 / f4 doesnt show anything out of the ordinary ...

any help? please!!

thx, and best rgds :( -pogbobo

here are some findings after playing around some more ... (although 2.6.2.2 is still not installing)

I dl'd 2.8.0 RC, burned it, and went to install it ... not only did it install just fine, all hardware compatible, it did shed some light on a possible reason why 2.6 wasnt working ... here is the error that was plastered at the top of the blue install screen after I chose my keyboard (2.8 mind you)

"ERROR: pdc: wrong # of devices in RAID set "pdc_bddgcijfa" [1/2] on /dev/hdc"

and then it hit me, the target hd used to be part of a raid1 setup, and i guess after all these years of using it as an XP disk, that raid info was still hiding somewhere in the MBR or first few sectors .... lame

so, after installing 2.8 just fine (using boot:advanced, choosing the option to remove all partition data and create default scheme), I pop in my handy UBCD4WIN in an attempt to actually clean out the MBR and all partition info ... well, I used Mbrfix and MBRWizzard but with no luck, the first said it cleaned the MBR just fine while the later just crapped out with an error ....

so in summary, 2.8 works, but I want 2.6, which still wont install .... any ideas on how to get that raid info purged out of the disk for good?

thx!

so, after a few days of pulling my hair out :) I seem to have found a solution, at least for my issues ...

as I stated in my previous post, there seems to be something about RAID metadata / signature information being left on the hard drive (from a previous hw raid setup i had) that just pisses off the TB install ... i tried cleaning the MBR because I though that was were the problem lied, well that was incorrect ... it turns out that the metadata / signature can be anywhere on the drive ... so, the catch-all solution was to write zeros to the entire disk, and boom, problem solved ... 2.6 installs just fine, and 2.8 doesnt have the weird error at the top :)

I used a knoppix 5.3 live dvd and entered the following command ... but be warned, this will zero out the taget disk, so be carful ...

dd_rescue -w /dev/hdx

x = a,b,c or whatever your drive is

Here's a good fix, hopefully for all of this. I'm posting this on this thread, even though there's a new thread...since i didnt find the latter even though I looked through this site a lot....

I just found a fix for the issue as described at the beginning of this thread...dunno about RAID as I'm not using it. I thought the issue was because of the SSD Intell X25 drive I'm using...anyway, no problems to install the latest Ubuntu. But CentOS would hang at "Loading ahci driver" and eventually report "no drive found".

so HUGE PROPS to "fakemoth" over at: http://www.centos.org/modules/newbb/viewtopic.php?topic_id=15720&...

who wrote, "To install CentOS (or others) with AHCI enabled in BIOS:
-start the drill
-when you get to the first screen DON't hit Enter cause we have to give a parameter to the kernel
-type: linux pci=nomsi boot and Enter

It always works."

Works for me. I've been trying for a week now to solve this install. I typed this single line:

linux pci=nomsi boot

and install just completed! :-)