10 More Discussions You Might Find Interesting
1. Solaris
Hello all;
We have a SunFire V240 with three disks that were part of a metadb. One of those disks, the boot disk, experienced a horrible death Monday night and we're now trying to recover from that.
I know very little about metadb stuff, so please be gentle with me...
I'm assuming... (3 Replies)
Discussion started by: Handon
3 Replies
2. SCO
Hi,
I need help, one of our servers is down, and all data in that Server are critical:
The Scenario:
-------------
I have a computer with the next architecture:
architecture=IA32
bus_types=PCI2.10,ISA,PnP1.0
hostname=xxx.xxx.xxx (the real name is hidden)
hw_provider=Generic AT... (2 Replies)
Discussion started by: danilosevilla
2 Replies
3. Solaris
I am supporting a server running Solaris 4.3.The server is not having ddicated system administrator and i am requred to do minor maintenance tasks.l From few days, The server is automatically crashing and i just reboot the server. I also do not see anything in /var/adm/messages that can give an... (1 Reply)
Discussion started by: asalman.qazi
1 Replies
4. Solaris
Hi all,
I want to put a local disk on a Sun Cluster node but scconf command explodes :eek:
My system:
* two node cluster on two VMWare virtual machines
* Solaris 10 SunOS 5.10 Generic_141415-05 i86pc i386 i86pc
* Sun Cluster 3.2 u2
* Veritas Volume manager
The situation... (2 Replies)
Discussion started by: gxmsgx
2 Replies
5. Solaris
hi ,
i have machine that is crashed
how i can enable core dump file & how can i find it ? :confused: (4 Replies)
Discussion started by: lid-j-one
4 Replies
6. Solaris
Can anyone of you help me in enabling crash dump on Solaris 5.5.1 (1 Reply)
Discussion started by: csreenivas
1 Replies
7. Solaris
I installed solaris 5.9 on a machine SunBlade100 512Mo of RAM and 18Go of hard disc, after the installation I have remark that the machine starts again after 10mn with 15 mn of walk, I downloaded and installed the last patchs on the machine but the poblème persists. someone can help me and tell me... (0 Replies)
Discussion started by: toufik
0 Replies
8. HP-UX
Hi,
First of all, thanks for your help. I have downloaded freeBSD to study unix
I'm here again 'cause my HP-UX 10.x has "broken". It raise this error:
----------------------------------------------------
Diagnostic System warning:
= 0x1f005000 is POWERFAILED The diagnostic logging... (1 Reply)
Discussion started by: efrenba
1 Replies
9. Filesystems, Disks and Memory
I need some expert help, and would appreciate any feedback on the following problem:
After power outage the machine didn't allow the login. When we checked it , it looked like both disks were completely empty. Luckily, we have a backup machine, and we can restore the disks.
Has anybody seen... (3 Replies)
Discussion started by: DGoubine
3 Replies
10. UNIX for Dummies Questions & Answers
We have a Unix system with two hard drives. When we tried to access it this morning it had frozen, and when we rebooted it couldn't find one of the hard drives, and had quite a few bad data blocks on the other. Of course we fear the worst: that the drive is completely gone, but were not sure how... (1 Reply)
Discussion started by: Kgatliffe
1 Replies
LVMCREATE_INITRD(8) System Manager's Manual LVMCREATE_INITRD(8)
NAME
lvmcreate_initrd - create an initial ramdisk to boot with root on a logical volume
SYNOPSIS
lvcmreate_initrd [-i|--iop IOP] [-D|--devfs] [-v|--verbose] [-V|--version] [KernelVersion]
DESCRIPTION
lvmcreate_initrd creates a new compressed initial ramdisk /boot/initrd-lvm-KernelVersion.gz. The initial ramdisk contains all necessary
binaries, shared libraries and a linuxrc file to switch to a logical volume based root filesystem. To build an initial ramdisk for a ker-
nel other than the one currently running add the KernelVersion parameter (e.g. lvmcreate_initrd 2.3.25) on the command line.
OPTIONS
-h, --help
Print a usage message on standard output and exit.
-i, --iop IOP
Specify the LVM kernel IOP version if you are using wrapper scripts in /sbin. This is only necessary if you are building an initrd
for a kernel other than the current one, and you have tools for multiple IOP versions installed.
-D, --devfs
Force the use of devfs when building the initrd. lvmcreate_initrd by default probes for devfs by checking for /dev/.devfsd, which
is created by devfsd when it is running. This option will force lvmcreate_initrd to do this, when devfsd is not currently running.
-v, --verbose
Gives verbose runtime information about the creation of the ram disk image.
-V, --version
Print version number and exit successfully.
The necessary actions to change your system into a "root on logical volume" one are:
Create a small (~20MB) partition which is BIOS reachable to hold the /boot filesystem (if you already have a small partition based root
filesystem this can be used instead). If you want to boot standalone from this partition in an emergency, copy all the necessary binaries
and libraries to that filesystem as well and create a corresponding /etc/lilo.conf entry. In order to be able to edit lilo.conf when booted
standalone, you should move /etc/lilo.conf to /boot/lilo.conf and create a symbolic link to it in /etc. This is not needed if you have a
boot/root floppy which contains the LVM binaries and the library.
Create all logical volumes you need (for root, usr, opt etc.), create filesystems in them, mount them and transfer all files from the par-
tition based filesystems into the logical volume based ones.
Set up your /etc/lilo.conf with a boot configuration like:
image = /boot/vmlinuz
initrd = /boot/initrd-lvm-KernelVersion.gz
root = /dev/YourVG/YourRootLV
label = rootonlv
append = 'ramdisk_size=8192'
Replace YourVG and YourRootLV by your actual volume group and root logical volume names. You also need to change the ramdisk size to be at
least as large as that reported while lvmcreate_initrd is run. If you have enough memory, it is OK if the ramdisk size in /etc/lilo.conf
is larger than what lvmcreate_initrd reports (the memory is freed after booting). Also, your /etc/fstab in the root logical volume should
contain entries for the root LV, and the boot partition, along with any other LVs you have configured:
/dev/YourVG/YourRootLV / ext2 defaults 0 1
/dev/YourBootPartition /boot ext2 defaults 0 2
/dev/YourVG/YourUsrLV /usr ext2 defaults 0 3
/dev/YourVG/YourOptLV /opt ext2 defaults 0 4
etc.
You can use other supported filesystem types as well (e.g. reiserfs) if you have support for those in your kernel. Run lilo, reboot and
try...
The partitions containing the former /usr, /opt etc. filesystems can now be used as physical volumes. Use pvcreate(8) to turn them into
PVs and then use vgextend(8) to add them to e.g. YourVG.
DIAGNOSTICS
lvcmcreate_initrd returns 0 for success. 1 is returned in all other cases.
SEE ALSO
lilo(8), lilo.conf(5), fstab(5), lvm(8), pvcreate(8), vgextend(8), lvcreate(8), vgcreate(8)
AUTHOR
Heinz Mauelshagen <Linux-LVM@Sistina.com>
Heinz Mauelshagen LVM TOOLS LVMCREATE_INITRD(8)