jim80.net

Welcome to my scratchpad. See links for some excellent write ups.

This article is based on Solaris 11 Express and VMWare ESXi 4.1.

Step 0: Build your hardware and install Solaris.

HCL: http://www.oracle.com/webfolder/technetwork/hcl/index.html

Step 1: Mirror the rpool

zpool status #figure out what’s already allocated

cfgadm -s "select=type(disk)" #list disks

fdisk -W - [rootdisk2] #check

fdisk -B [rootdisk2] #apply a default Solaris partition to the disk

fdisk -W - [rootdisk2] #check again

prtvtoc /dev/rdsk/[rootdisk1] | fmthard -s - /dev/rdsk/[rootdisk2] #slice and dice! You must define slices for rpools, so this step mirrors the original disk’s slice-age to the second root disk

zpool attach -f rpool [rootdisk1] [rootdisk2] #attach, not add. Attach = mirror. Add = stripe.

zpool status #check

installgrub /boot/grub/stage1 /boot/grub/stage2 /dev/rdsk/[rootdisk2] #install grub onto second boot disk

zpool status #check again

see: http://constantin.glez.de/blog/2011/03/how-set-zfs-root-pool-mirror-oracle-solaris-11-express

For more on rpool manipulation, namely on how to shrink an rpool, see: http://blogs.oracle.com/mock/entry/how_to_shrink_a_mirrored

Step 2: Configure networking

ifconfig -a #cheat and take advantage of nwam’s device lists

svcadm disable network/physical:nwam #disable nwam

svcadm enable network/physical:default #then enable the default config tools

ipadm create-if e1000g0 #or whatever your nic is

ipadm show-if #check

ipadm create-addr -T static -a mg.mt.ip.addy/prefix e1000g0/v4 #create the address

ipadm show-addr #confirm

#link aggregate group (LACP)

dladm create-aggr -l [nic2] -l [nic 3] #create link aggregate group:

Then the same ipadm commands

ipadm create-if aggr1 #configure aggr1 for persistence

ipadm show-if #confirm it was added, should show down.

ipadm create-addr -T static -a 1.2.3.4/24 aggr1/v4 #create the address

ipadm show-addr #confirm it was added, show-if should now show “OK”

netstat -rn #check routing tables

route -p add default 1.2.3.1 #add persistent default route

vi /etc/resolv.conf #configure nameservers

vi /etc/hosts #configure name

vi /etc/nsswitch.conf #enable dns lookups for “hosts” line

svcsadm restart ssh #restart ssh after modifying networking

ping 4.2.2.1 #ping Level3′s DNS resolvers to confirm routing

reboot #reboot to confirm persistency

Step 3: Do your AD thing. I know you love it.

read this: http://blog.scottlowe.org/2006/08/15/solaris-10-and-active-directory-integration/

also: http://download.oracle.com/docs/cd/E19963-01/html/821-1449/index.html

Step 4: Configure ZFS

format #to present list of disks. exit out, leaving the list in screen buffer

or: cfgadm -s "select=type(disk)"

zpool create tank mirror c7t0d0 c7t1d0 mirror c7t2d0 c7t3d0 cache c7t4d0 #create your main zpool. This is 4 drives in raid 10 and a read cache. use “log c#t#d#” to add a ZIL device

zfs set dedup=on tank #enable dedup

zfs set compression=on tank #enable compression.

Note that changes made to the pool will apply to its children when they are created.

zfs create tank/VMOS #create your ESXi share

zfs set sharenfs=none,nosuid,root=1.2.3.xx:1.2.3.xy tank/VMOS #share with 1.2.3.xx and 1.2.3.xy

From esxi 4.1:

esxcfg-nas -a -o IP.Or.Host.Name -s /tank/VMOS [esxiDatastoreName]# http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1005057

ZFS /Solaris links

http://lildude.co.uk/zfs-cheatsheet

http://download.oracle.com/docs/cd/E19963-01/index.html

iSCSI - Solaris 11 = use COMSTAR.

http://wikis.sun.com/display/OpenSolarisInfo200906/Configuring+an+iSCSI+Storage+Array+With+COMSTAR+%28Task+Map%29

Tags: ESXi 4.1, mirror rpool, NFS, Solaris 11 Express, ZFS, zpool

This entry was posted on Sunday, August 7th, 2011 at 18:36 by jim and is filed under AD, Infrastructure, Storage, Virtualization, Windows. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.