After a few weeks of testing and experimenting, finally found out the issue of the setup:some repos were not enable by default
Now, I had figure the issue and put the changes into cfengine (yes I know is old school) and will also include in ansible playbook (WIP).
I don’t know how to cook except fried egg, and boil water (for instance noodles and 3 in 1 coffee/tea).
I don’t know how to play any music, can’t even sing without hurting others’ eardrums.
I don’t know how to sew – can’t even sew back a button on the shirt.
I don’t know how to …
What am I good in doing? Maybe only 1 thing I am really good in – being myself!
All of the sudden, I felt like the time had halt, and nothing seem to move.
This is just a small stopover, and I shall start again soon.
I am just here to deliver my message
The important part is the learning along the journey, winning is just the icing on top of it!
It seem like CentOS7/RHEL7 is slightly different from CentOS6/RHEL6, when trying to install from USB with multiple boot images (eg YUMI).
CentOS7/RHEL7 require “inst.stage2″^^1 to be specific, and the default from the ISO file is “inst.stage2=hd:LABEL=CentOS\x207\x20x86_64” or “inst.stage2=hd:LABEL=RHEL-7.3\x20x86_64”. This will need to be amend to reflect the correct LABEL of the usb.
How to find out the LABEL for your USB? You can refer to here. But to save your time (since you found this instead of that):
1. Boot into Linux
2. Assume your USB is on /dev/sda, and the first partition contain your multiboot:
Example of the label on my USB is “MULTIBOOT”, therefore I need to change the “inst.stage2” to:
which will load the stage2 image from /dev/sda1/multiboot/CentOS-7
^^1 – This specifies the location to fetch only the installer runtime image; packages will be ignored.
^^2 – More info on the Boot options from here.
I just got a new Dell Precision T3420 for testing RHEL6 installation on it.I had struggle for a few days as the nvme driver (0.10) that’s in kernel-2.6.32 is unable to handle the Interrupt remapping done by the BIOS, and keep throwing up error such as:
ERR kernel:dmar: DRHD: handling fault status reg 2
ERR kernel:dmar: INTR-REMAP: Request device [[02:00.0] fault index 26
ERR kernel:INTR-REMAP:[fault reason 38] Blocked an interrupt request due to source-id verification failure
RHEL6 still able to see the NVMe PCI SSD, but installation is not smooth and even booting up after installation the errors are still there.
I was seaching newer nvme driver for kernel-2.6.32 but can’t find any usable. RHEL7 with the newer nvme driver have no issue of installation.
Luckily, a colleague had done it previously told me that this is cause by interrupt remapping, and I should turn it off during installation and during boot.
Here is the extra option to be put in:
And here is an example of how grub.conf should looked like:
title Red Hat Enterprise Linux 6 (2.6.32-642.el6.x86_64)
kernel /vmlinuz-2.6.32-642.el6.x86_64 ro root=/dev/mapper/vg_t3420test-lv_root rd_NO_LUKS KEYBOARDTYPE=pc KEYTABLE=us LANG=en_US.UTF-8 rd_LVM_LV=vg_t3420test/lv_swap rd_NO_MD rd_LVM_LV=vg_t3420test/lv_root SYSFONT=latarcyrheb-sun16 rd_NO_DM rhgb quiet nointremap
I wasn’t able to find anything about this online, so I thought I should write this down in my blog.
Hope this help someone else who face the same problem.