ESXi 4.1.0 Update 1 : How to mount a remote ESXi’s vmfs from Ubuntu

I referred to the following URL.

http://thebsdbox.co.uk/?p=224

Ubuntu 11.04

VMware ESXi 4.1.0 build-348481

             

              mount ESXi’s vmfs

Ubuntu ---------------------------- ESXi

[ preparation ]

[ on Ubuntu ]

install sshfs

# apt-get install sshfs

To mount vmfs with sshfs from ubuntu , sftp is needed on a remote machine(ESXi).

copy sftp-server binary file to ESXi’s /sbin/directory.

on Ubuntu , download sftp-server.tar.gz , extract tar.gz , copy sftp-server file to ESXi’s /sbin directory.

# wget http://thebsdbox.co.uk/wp-content/uploads/2010/08/sftp-server.tar.gz # tar xzvf sftp-server.tar.gz sftp-server # file sftp-server sftp-server: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), statically linked, for GNU/Linux 2.2.5, not stripped # scp sftp-server root@vmware:/sbin/

confirm there’s sftp-server file under ESXi’s /sbin directory.

[ on ESXi ]

~ # vmware -lv VMware ESXi 4.1.0 build-348481 VMware ESXi 4.1.0 Update 1~ # ls /sbin/sftp-server /sbin/sftp-server

[ mount a remote ESXi’s vmfs from Ubuntu ]

[ on Ubuntu ]

# sshfs -o idmap=user root@vmware:/ /root/tmp_dir root@vmware's password: # cd /root/tmp_dir/ # ls altbootbank  dev  lib64      opt            sbin     tmp  vmfs bin          etc  local.tgz  proc           scratch  usr  vmimages bootbank     lib  locker     productLocker  store    var  vmupgrade

unmount

# fusermount -u /root/tmp_dir

[ tips ]

Unfortunately , when rebooting ESXi , /sbin/sftp-server file will be removed.

To solve this , copy sftp-server file to ESXi’s oem.tgz file.

[ on Ubuntu ]

# scp root@vmware:/bootbank/oem.tgz . root@vmware's password: oem.tgz                                       100%  928KB 927.8KB/s   00:00 # cd zzz/ # cp /root/oem.tgz . # pwd /root/zzz # ls oem.tgz # tar xzvf oem.tgz # rm oem.tgz # ls etc  lib  oem.tgz  sbin  usr  var # cp /root/sftp-server sbin/ # tar czvf oem.tgz * etc/ etc/cim/ etc/vmware/ etc/vmware/init/ etc/vmware/init/manifests/ etc/vmware/init/manifests/Neterion-vxge.mf etc/vmware/init/manifests/LSI-3w-9xxx.mf etc/vmware/pciid/ etc/vmware/pciid/vxge.xml etc/vmware/pciid/3w-9xxx.xml etc/ipmi/ lib/ lib/cim/ sbin/ sbin/sftp-server usr/ usr/lib/ usr/lib/vmware/ usr/lib/vmware/vmkmod/ usr/lib/vmware/vmkmod/vxge.o usr/lib/vmware/vmkmod/3w-9xxx.o usr/lib/ext/ usr/lib/pycim/ var/ var/db/ var/db/esxupdate/ var/db/esxupdate/oem-pkgdb.tgz var/lib/ var/lib/sfcb/ var/lib/sfcb/registration/

copy oem.tgz to ESXi’s /bootbank directory

[ on Ubuntu ]

# scp oem.tgz root@vmware:/bootbank/ root@vmware's password: oem.tgz                                       100% 2002KB   2.0MB/s   00:00

[ on ESXi ]

After rebooting ESXi , there’s sftp-server file under /sbin directory.

Tech Support Mode may be disabled by an administrative user. Please consult the ESXi Configuration Guide for additional important information. ~ # ls /sbin/sftp-server /sbin/sftp-server

ESXi 4.1.0 Update 1: How to make a clone VM via CLI

~ # vmware -v VMware ESXi 4.1.0 build-348481

create a directory for a clone VM

~ # mkdir /vmfs/volumes/datastore1/CLONE

make a clone VM with vmkfstools command.

vmkfstools –i “/parh/original VM.vmdk” “/path/clone VM.vmdk”

~#vmkfstools -i /vmfs/volumes/datastore1/ORIGINAL/ORIGINAL.vmdk /vmfs/volumes/datastore1/CLONE/CLONE.vmdk Destination disk format: VMFS zeroedthick Cloning disk '/vmfs/volumes/datastore1/ORIGINAL/ORIGINAL.vmdk'... Clone: 100% done. ~ #

copy original VM’s vmx file to clone VM’s directory.

# cp ../ORIGINAL/ORIGINAL.vmx CLONE.vmx

modify vmx file

# cat CLONE.vmx | grep -v "ethernet0.generatedAddress = " | grep -v "ethernet0.generatedAddressOffset = " | grep -v "uuid.location = " | grep -v "uuid.bios = " | sed 's/'"ORIGINAL"'/'"CLONE"'/g' > vmx.tmp | echo 'uuid.action = "keep"' >> vmx.tmp # cp vmx.tmp CLONE.vmx

register a clone VM

please note that you need to specify full path.

# /bin/vim-cmd solo/registervm /vmfs/volumes/datastore1/CLONE/CLONE.vmx

ESXi 4.1 update 1: How to change language environment of vSphere Client

vSphere Client Version : 4.1.0 Build 345043

VMware ESXi Version : 4.1.0 Build 348481

I installed vSphere client with Japanese language environment , but I want to change it in English.

There are two ways

- method 1

open command prompt and start vSphere client ( VpxClient.exe ) with “--locale en_US” option.

>"C:\ProgramFiles\VMware\Infrastructure\VirtualInfrastructureClient\Launcher\VpxClient.exe" --locale en_US

- method 2

open VMware vSphere Client property and add “--local en_US”.

Language environment has changed in English.

VMware ESXi 4.1.0 Update 1 : CPU Hot plug

I referred to the following URL.
http://communities.vmware.com/docs/DOC-10493

[ software info ]

ESXi

~ # vmware -l VMware ESXi 4.1.0 Update 1

VM is Scientific Linux6 (SL6) 32bit. I assigned 1 vCPU.

[root@sl6-1 ~]# cat /etc/redhat-release Scientific Linux release 6.0 (Carbon) [root@sl6-1 ~]# uname -r 2.6.32-131.6.1.el6.i686

[ procedures ]

Step1

power off VM

Step2

on vSphere client , enable CPU hot plug to that VM.

Step3

power on VM

Step4

add vCPUs

on vSphere client , VM -> edit -> Hardware -> CPU
I changed # of vCPUs from 1 to 4.


dmesg

[root@sl6-1 ~]# dmesg processor LNXCPU:03: registered as cooling_device3 SMP alternatives: switching to SMP code Booting Node 0 Processor 3 APIC 0x3 Initializing CPU#3 mce: CPU supports 0 MCE banks x86 PAT enabled: cpu 3, old 0x0, new 0x7010600070106 Skipped synchronization checks as TSC is reliable. Booting Node 0 Processor 2 APIC 0x2 Initializing CPU#2 mce: CPU supports 0 MCE banks x86 PAT enabled: cpu 2, old 0x0, new 0x7010600070106 Skipped synchronization checks as TSC is reliable. Booting Node 0 Processor 1 APIC 0x1 Initializing CPU#1 mce: CPU supports 0 MCE banks x86 PAT enabled: cpu 1, old 0x0, new 0x7010600070106 Skipped synchronization checks as TSC is reliable.

# of vCPU truns 4.

[root@sl6-1 ~]# cat /proc/cpuinfo | grep "model name" | wc -l 4

On SL6 32bit , I don’t need to do anything to activate hot-plugged vCPUs.
Just only change # of vCPUs via vSphere client.

change # of vCPUs from 4 to 8.

# of vCPU turns 8.

[root@sl6-1 ~]# cat /proc/cpuinfo | grep "model name" | wc -l 8

dmesg

processor LNXCPU:07: registered as cooling_device7 Booting Node 0 Processor 6 APIC 0x6 Initializing CPU#6 mce: CPU supports 0 MCE banks x86 PAT enabled: cpu 6, old 0x0, new 0x7010600070106 Skipped synchronization checks as TSC is reliable. Booting Node 0 Processor 5 APIC 0x5 Initializing CPU#5 mce: CPU supports 0 MCE banks x86 PAT enabled: cpu 5, old 0x0, new 0x7010600070106 Skipped synchronization checks as TSC is reliable. Booting Node 0 Processor 4 APIC 0x4 Initializing CPU#4 mce: CPU supports 0 MCE banks x86 PAT enabled: cpu 4, old 0x0, new 0x7010600070106 Skipped synchronization checks as TSC is reliable. Booting Node 0 Processor 7 APIC 0x7 Initializing CPU#7 mce: CPU supports 0 MCE banks x86 PAT enabled: cpu 7, old 0x0, new 0x7010600070106 Skipped synchronization checks as TSC is reliable.

[root@sl6-1 cpu]# pwd /sys/devices/system/cpu [root@sl6-1 cpu]# ls cpu0  cpu2  cpu4  cpu6  cpufreq  kernel_max  online    present cpu1  cpu3  cpu5  cpu7  cpuidle  offline     possible [root@sl6-1 cpu]# [root@sl6-1 cpu]# cat online 0-7