VM Management

# powershell script for VM mass management. Requires a .csv file containing the list of VMs and a name for the snapshot/ If no snapshot name is given, "Snapshot_1" is used.

 #load powercli if needed
if (!(Get-Module -Name VMware.VimAutomation.Core) -and (Get-Module -ListAvailable -Name VMware.VimAutomation.Core)) {
    Write-Output "loading the VMware Core Module..."
    if (!(Import-Module -Name VMware.VimAutomation.Core -ErrorAction SilentlyContinue)) {
        # Error out if loading fails
        Write-Error "`nERROR: Cannot load the VMware Module. Is the PowerCLI installed?"
     }
    $Loaded = $True
    }
 #   elseif (!(Get-Module -Name VMware.VimAutomation.Core -ErrorAction SilentlyContinue) -and !(Get-Module -Name VMware.VimAutomation.Core) -and ($Loaded -ne $True)) {
 #       Write-Output "loading the VMware Core Snapin..."
 #    if (!(Add-PSSnapin -PassThru VMware.VimAutomation.Core -ErrorAction SilentlyContinue)) {
 #    # Error out if loading fails
 #    Write-Error "`nERROR: Cannot load the VMware Snapin or Module. Is the PowerCLI installed?"
 #    }
 #   }

# Define vmConfigSpec params
$vmConfigSpec = New-Object VMware.Vim.VirtualMachineConfigSpec
$vmConfigSpec.Tools = New-Object VMware.Vim.ToolsConfigInfo
$vmConfigSpec.Tools.ToolsUpgradePolicy = "UpgradeAtPowerCycle"

# Get the command-line params
 $command = $args[0]
 $list = $args[1]
 $server = $args[2]

# define default params
if (!$list) { $list = "vm_mgmt.csv"}
if (!$server) { $server = "default.vCenter.domain.tld"}

# Start processing the command
  switch ($command) {
    default { $myname = $MyInvocation.MyCommand.Definition
    echo "`nERROR! Usage:"
    echo "$myname command [list] [server] "
    echo "`ncommand is one of the following: viewsnap, takesnap, delsnap, revertsnap, hwupd, vmtoolsupd, vmoff, vmon."
    echo " list is a .csv file containing 'VM_Name,Snapshot_name'. If no list provided, 'vm_mgmt.csv' will be used."
    echo " server is the name of the server connecting to. If no server is provided, 'default.vCenter.domain.tld' will be used.`n"
    }
 
  "viewsnap" {
    ### View Snapshot
      Connect-VIServer -Server $server -Protocol https
      import-csv $list | ForEach-Object {
      $_.VM_Name
      $_.Snapshot_name
      if (!$_.Snapshot_name) { $_.Snapshot_name = "Snaphot_1"}
    echo "`nView Snaphots:`n"
      get-vm -Name $_.VM_Name | get-snapshot
  } }

  "takesnap" {   
    ### Take Snapshot
      Connect-VIServer -Server $server -Protocol https
      import-csv $list | ForEach-Object {
      $_.VM_Name
      $_.Snapshot_name
      if (!$_.Snapshot_name) { $_.Snapshot_name = "Snaphot_1"}
    echo "`nTaking Snapshots`n"
      get-vm -Name $_.VM_Name | New-Snapshot -Name $_.Snapshot_1 -Quiesce -Memory
  } }

  "delsnap" {   
    ### Delete Snapshot
    Connect-VIServer -Server $server -Protocol https
    import-csv $list | ForEach-Object {
    $_.VM_Name
    $_.Snapshot_name
      if (!$_.Snapshot_name) { $_.Snapshot_name = "Snaphot_1"}
   echo "`nDelete Snapshots`n"
    get-snapshot -name $_.Snapshot_1 -vm $_.VM_Name | remove-snapshot -confirm:$false
  } }

  "revertsnap" {
    ### Revert To Snapshot
      Connect-VIServer -Server $server -Protocol https
      import-csv $list | ForEach-Object {
      $_.VM_Name
      $_.Snapshot_name
    echo "`nReverting Snapshots. Confirmation is required for each restore.`n"
      set-vm -VM $_.VM_Name -Snapshot $_.Snapshot_1 -whatif
    # set-vm -VM $_.VM_Name -Snapshot $_.Snapshot_1 -confirm:$false
  } }

  "hwupd" { 
    # VM Hardware upgrade
      Connect-VIServer -Server $server -Protocol https
      import-csv $list | ForEach-Object {
      $_.VM_Name
      $_.Snapshot_name
      if (!$_.Snapshot_name) { $_.Snapshot_name = "Snaphot_1"}
    echo "`nVM Hardware upgrade to vmx-13`n"
      $do = New-Object -TypeName VMware.Vim.VirtualMachineConfigSpec
      $do.ScheduledHardwareUpgradeInfo = New-Object -TypeName VMware.Vim.ScheduledHardwareUpgradeInfo
      $do.ScheduledHardwareUpgradeInfo.UpgradePolicy = “always”
      $do.ScheduledHardwareUpgradeInfo.VersionKey = “vmx-13”
      $vm.ExtensionData.ReconfigVM_Task($do)
  } }

  "vmtoolsupd" {
    # VM Tools update
      Connect-VIServer -Server $server -Protocol https
      import-csv $list | ForEach-Object {
      $_.VM_Name
      $_.Snapshot_name
      if (!$_.Snapshot_name) { $_.Snapshot_name = "Snaphot_1"}
    echo "`nUpdating VM Tools`n"
      get-vm -Name $_.VM_Name | %{$_.Extensiondata.ReconfigVM($vmConfigSpec)}
  } }

  "vmoff" {
    # VM power off
      Connect-VIServer -Server $server -Protocol https
      import-csv $list | ForEach-Object {
      $_.VM_Name
      $_.Snapshot_name
      if (!$_.Snapshot_name) { $_.Snapshot_name = "Snaphot_1"}
    echo "`nTurning VMs OFF`n"
      $vm = Get-VM -Name $_.VM_Name | Shutdown-VMGuest -Confirm:$false
  } }

  "vmon" {
    # VM power on
      Connect-VIServer -Server $server -Protocol https
      import-csv $list | ForEach-Object {
      $_.VM_Name
      $_.Snapshot_name
      if (!$_.Snapshot_name) { $_.Snapshot_name = "Snaphot_1"}
    echo "`nTurning VMs ON`n"
      $vm = Get-VM -Name $_.VM_Name | Start-VM -Confirm:$false
  } }

}



-----------------------------------------------
type vm_mgmt.csv

VM_NAME,Snapshot_name
some-vm-name,Snapshot_342
another-vm-name,Snapshot_temp4

Chromium - Cast: "No Devices Found"

Starting chromium from the console you get "Component extension with id pkedcjkdefgpdelpbcmbmeomcjbeemfm not in whitelist and is not being loaded as a result."

- pkedcjkdefgpdelpbcmbmeomcjbeemfm is the Chrome Media Router, without it, you can go to chrome://flags and Enable the #load-media-router-component-extension as much as you want, no chromecast device will ever be detected.

My solution: 

1: copy pkedcjkdefgpdelpbcmbmeomcjbeemfm folder from a google-chrome profile ( ~/.config/google-chrome/Default/Extensions/pkedcjkdefgpdelpbcmbmeomcjbeemfm) to the chromium profile ~/.config/chromium/Default/Extensions/pkedcjkdefgpdelpbcmbmeomcjbeemfm

( If you copy from a windows installation, make sure the files are chmod 600, the folders are 700 and you are the right owner)

2: in Chromium, go to chrome://extensions/, enable "Developer Mode", then "Load Unpacked" and point to ~/.config/chromium/Default/Extensions/pkedcjkdefgpdelpbcmbmeomcjbeemfm

3: Now Chrome Media Router is loaded and the chromecast devices will be detected:

drakboot "INTERNAL ERROR: unknown device"

1: obviously, run strace drakboot > /tmp/trace 2>&1

2: less /tmp/trace, look for the missing device (in my case it was sdd):

 openat(AT_FDCWD, "/boot/grub2/install.sh", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 12

ioctl(12, TCGETS, 0xbfb7147c)           = -1 ENOTTY (Inappropriate ioctl for device)

_llseek(12, 0, [0], SEEK_CUR)           = 0

fstat64(12, {st_mode=S_IFREG|0755, st_size=22, ...}) = 0

read(12, "grub2-install /dev/sdd", 8192) = 22

read(12, "", 8192)                      = 0

3: edit /boot/grub2/install.sh and replace the wrong disk reference.

Push-button ON/OFF

In order to start the Peltier-based car freezer and to make sure that it turn off and stays OFF after the car stopped (a big Peltier is a huge battery drain!), I've made this 4-relay based circuit. No or "as little as possible" electronics in the car is the best approach - only the aerospace is harder for electronics than automotive - the best solution might involve some relays and, maybe, a couple of diodes, but stay away from amateurish electronics in the car! This is the diagram I draw on the back of a napkin (as usual) :)

13

All good!

RDP Disconnected! Error Code: 2308 Error Description: Socket closed

After a windows update, a couple of Windows 2016 Servers on AWS started rejecting the RDP connections.
mRemote was giving the error "RDP Disconnected! Error Code: 2308 Error Description: Socket closed". MS RDP is giving "This computer can't connect to the remote computer. Try connecting again. If the problem continues, contact the owner of the remote computer or your network administrator."

After a bit of tinkering, I found that the problem seems to be the RDP TLS and encryption level.

To solve it:

- remote connect Registry Editor to the affected server and change the DWORD 
HKLM\CurrentControlSet\Control\Terminal Server\WinStations\RDP-Tcp\SecurityLayer from "2" to "0" 


- remote connect Services.msc to the affected server and restart TermService

Windows could not complete the installation. To install windows on this computer restart the installation

SHIFT-F10 to bring up command prompt.
cd c:\windows\system32\oobe
msoobe.exe
Make a generic account and password. 
Hit finish (if it requests a product key and you have one, enter it now.  if OEM/No key required, just finish). 
Set time/date. 
Finish.

Extend LVM with a new disk

partprobe 
cat /proc/partitions - note the un-partitioned drive (ex: there is a /dev/xvdX but not /dev/xvdX1) 
cfdisk /dev/xvdX - create new partition, type 8e (LVM) 
pvcreate /dev/xvdX1 
vgdisplay  - note the VG name (in my case is "vg_bkp_vol_1")
vgextend vg_bkp_vol_1 /dev/xvdX1 
pvscan 
lvdisplay  - note the LV path (here "/dev/vg_bkp_vol_1/lv_bkp_vol_1")
lvextend /dev/vg_bkp_vol_1/lv_bkp_vol_1 /dev/xvdX1 
lvdisplay 
vgdisplay 
resize2fs /dev/vg_bkp_vol_1/lv_bkp_vol_1 
df 

12

Right on time!

Deploy .pfx cert embedded in script (a sort of 'cat << EOF' for windows)

@echo off
::
::  
:: In order to prepare the certificate please run
:: 'certutil -encode the_pfx_cert base_64_cert`
:: then paste the base_64_cert in the section below
:: Please note that the certificate password has to be given as start paramater to this script!
:: (eq: "cert-inst.bat S3cr3tPassw0rd")

:: If the cert was already installed, exit
REG QUERY HKCU\SOFTWARE\neXt /v CertInstalled
If %errorlevel%==0 goto :eof

:: define the temp name of the extracted cert
set extractedfile=%temp%\extract-%random%.txt

:: set the password needed to decode the cert
set certpasswd=%~1

:: separate the cert from this script
call:extractembedded embeddedfile %extractedfile%

:: process the extracted file
certutil -decode %extractedfile% %extractedfile%.pfx

certutil -f -user -p %certpasswd% -importpfx %extractedfile%.pfx

:: clean-up
::del %extractedfile% %extractedfile%.pfx

:: leave a trace in the registry, so the cert will not be installed again and again
REG ADD HKCU\SOFTWARE\neXt /v CertInstalled /t REG_DWORD /d 1

:: clean exit
exit /b

:: begin of the embed cert & extraction procedure
:: After the next line, please paste the "base_64_cert" created by certutil -encode
goto:embeddedfile
-----BEGIN CERTIFICATE-----
MIIMngIBAzCCDGQGCSqG
[...]
k05EzAQIFXJaGHOuxZcCAggA
-----END CERTIFICATE-----
:embeddedfile
:: before the previous line you can find the end of the "base_64_cert"

:: cert extraction procedure
:extractembedded
setlocal EnableDelayedExpansion
set embedbegin=goto:%~1
set embedend=:%~1
set embedcert=%~2
if exist %embedcert% del %embedcert%
set tmprndfile=%temp%\%random%.%random%
findstr /n ^^ "%~f0" > %tmprndfile%
call :seekembed < %tmprndfile%
del %tmprndfile%
exit /B
:seekembed
set oneline=:eof
set /P oneline=
if !oneline! == :eof goto nostart
set oneline=!oneline:*:=!
if not !oneline! == %embedbegin% goto seekembed
:getline
set oneline=:eof
set /P oneline=
if !oneline! == :eof goto nostop
set oneline=!oneline:*:=!
if !oneline! == %embedend% goto :eof
echo/!oneline!>> %embedcert%
goto getline
:nostart
echo Error finding start delimiter %embedbegin%
goto :eof
:nostop
echo Error finding stop delimiter %embedend%
goto :eof

Boot from grub2 rescue prompt

grub rescue> set prefix=(hd0,1)/boot/grub2
grub rescue> set root=(hd0,1)
grub rescue> insmod normal
grub rescue> normal
grub rescue> insmod linux
grub rescue> linux /boot/vmlinuz root=/dev/sda1
grub rescue> initrd /boot/initrd.img
grub rescue> boot

SCO SYSV on a HP server. Install notes

Phone extension that rings multiple external numbers (CCME Asterisk)

Skip to end of metadata

o to start of metadata

In this example we want to ring all the external numbers of a few people only by dialing extension 3331.

1: Connect to CCME and create a new dial-peer:

dial-peer voice 3331 voip
 description External Emergency Responders
 destination-pattern 3331$
 session protocol sipv2
 session target ipv4:192.168.0.XXX #(the Asterisk box)
 dtmf-relay rtp-nte cisco-rtp
 codec g711ulaw
 no vad

2: Connect to the asterisk server and add the extension in the default [incoming_context] in /etc/asterisk/extensions.conf: (192.168.0.YYY is the outgoing trunk)

exten => 3331,1,Dial(SIP/5145555555@192.168.0.YYY&SIP/51455555519@192.168.0.YYY&SIP/5145555552@192.168.0.YYY&SIP/5145555553@192.168.0.YYY&SIP/5145555554@192.168.0.YYY&SIP/5145555555@192.168.0.YYY)
exten => 3331,n,Hangup

3: Reload asterisk config and test the extension 

Limit number of unix logins

cat .profile

#!/bin/sh
limit=3

session=`ps -ef | grep '\-sh' | grep $USER | grep -v grep`

number=`echo $session | wc -l`

if [ $number -ge $limit ]; then
echo "No more logins / Il n'y a plus de login. You are already logged as:
$session "

sleep 5
exit 0
fi

# to timeout after 15min of inactivity and forbid users to change the tmout:
echo "TMOUT=900
readonly TMOUT
export TMOUT" > /etc/profile.d/tmout.sh && chmox +x /etc/profile.d/tmout.sh

Forward all emails from one sendmail to another

- in sendmail.cf change Djdomain.tld to something like Dj_subdomain.domain.tld Do the same thing for DMdomain.tld
- in the same sendmail.cf look for Fw or Fw-o and check the file it refers to (usually /usr/lib/mail/local-host-names); in that file remove the line stating "domain.tld" (in fact, remove everything, leave just _subdomain.domain.tld there), that way sendmail will know that it is not the default destination for the @domain.tld

- if necessary, create /home/$username/.forward (chmod 600, chown $username:$group) and put the right info in it

Install mc on sco sysV unix

curl http://www.aljex.com/bkw/sco/mc.tar.bz2 |bzcat |tar xvf -
(there is a copy of that archive in my /update directory)

in .profile add PATH=$PATH:/usr/local/bin:/usr/local/sbin

Dell Latitude Keyboard light

The backlit keys interface seems to be broken, so you need to mask the controlling service: systemctl mask systemd-backlight@leds\:dell\:\:kbd_backlight.service
then go to BIOS and and change the keyboard led timeout from 10s to 5s for AC Power and Battery mode - "For some reason if the kbd_backlight service is not masked, it somehow forces the bios led timeout back to 10s, and then the keyboard led always remains lit."

TS - FR keyboard by default

Keyboard
Data collected on: 4/21/2015 1:44:16 PM

General

Details

Domain	domain.local
Owner	DOMAIN\Admins du domaine
Created	4/21/2015 1:32:00 PM
Modified	4/21/2015 1:42:44 PM
User Revisions	18 (AD), 18 (sysvol)
Computer Revisions	1 (AD), 1 (sysvol)
Unique ID	{0E240A4C-8A26-4761-8907-DB164F024AFC}
GPO Status	Enabled

Links

Location	Enforced	Link Status	Path
TS	No	Enabled	domain.local/Member Servers/TS

This list only includes links in the domain of the GPO.

Security Filtering

The settings in this GPO can only apply to the following groups, users, and computers:

Name

NT AUTHORITY\Authenticated Users

Delegation

These groups and users have the specified permission for this GPO

Name	Allowed Permissions	Inherited
NT AUTHORITY\Authenticated Users	Read (from Security Filtering)	No
NT AUTHORITY\ENTERPRISE DOMAIN CONTROLLERS	Read	No
NT AUTHORITY\SYSTEM	Edit settings, delete, modify security	No
		No

Computer Configuration (Enabled)

Policies

Administrative Templates

Policy definitions (ADMX files) retrieved from the local machine.

System/Group Policy

Policy	Setting	Comment
User Group Policy loopback processing mode	Enabled
Mode: Merge

User Configuration (Enabled)

Preferences

Windows Settings

Registry

2 (Order: 1)

General

Action

Update

Properties

Hive	HKEY_CURRENT_USER
Key path	Keyboard Layout\Preload
Value name	2
Value type	REG_SZ
Value data	00000409

Common

Options

Stop processing items on this extension if an error occurs on this item	No
Run in logged-on user's security context (user policy option)	No
Remove this item when it is no longer applied	No
Apply once and do not reapply	No

1 (Order: 2)

General

Action

Update

Properties

Hive	HKEY_CURRENT_USER
Key path	Keyboard Layout\Preload
Value name	1
Value type	REG_SZ
Value data	00000c0c

Common

Options

Stop processing items on this extension if an error occurs on this item	No
Run in logged-on user's security context (user policy option)	No
Remove this item when it is no longer applied	No
Apply once and do not reapply	No

ShowStatus (Order: 3)

General

Action

Update

Properties

Hive	HKEY_CURRENT_USER
Key path	Software\Microsoft\CTF\LangBar
Value name	ShowStatus
Value type	REG_DWORD
Value data	0x4 (4)

Common

Options

Stop processing items on this extension if an error occurs on this item	No
Run in logged-on user's security context (user policy option)	No
Remove this item when it is no longer applied	No
Apply once and do not reapply	No

Label (Order: 4)

General

Action

Update

Properties

Hive	HKEY_CURRENT_USER
Key path	Software\Microsoft\CTF\LangBar
Value name	Label
Value type	REG_DWORD
Value data	0x0 (0)

Common

Options

Stop processing items on this extension if an error occurs on this item	No
Run in logged-on user's security context (user policy option)	No
Remove this item when it is no longer applied	No
Apply once and do not reapply	No

AutoAdjustDeskBand (Order: 5)

General

Action

Update

Properties

Hive	HKEY_CURRENT_USER
Key path	Software\Microsoft\CTF\MSUTB
Value name	AutoAdjustDeskBand
Value type	REG_DWORD
Value data	0x0 (0)

Common

Options

Stop processing items on this extension if an error occurs on this item	No
Run in logged-on user's security context (user policy option)	No
Remove this item when it is no longer applied	No
Apply once and do not reapply	No

SCO SysV Cannot enable remote printing (mkdev rlp) after the installation of an Update Pack 2/3 or a Maintenance Pack

"Unable to get complete data for 'Printer Manager on ...
Failed to retrieve list of remote printers.    
The TCP/IP rlp remote printing package is not installed...."
According to the documentation of Update Pack x and Maintenance
          Pack x, "mkdev rlp" MUST be run before applying Update Pack x or
          when including CUPS as part of the installation of the Maintenance
          Pack as detailed in:
         
          CUPS and Remote Printing (LPD)
         
          Although CUPS supports LPD as both a server and a client, the CUPS
          LPD server implementation does not support access control (based on
          the settings in the /etc/hosts.equiv and /etc/hosts.lpd files).
          If your setup requires the use of the standard LPD, or you wish to
          use access control, do not install CUPS.
WARNING  If you have never run mkdev rlp and you wish to do
          so, the CUPS package must be removed before running mkdev rlp
          and then reinstalled after remote printing is configured.
on a system-wide basis in /etc/default/lpd. The default entry is
            for the SYSV print system:

            PRINT_SYSTEM=SYSV


NOTE:
          If you already had run "mkdev rlp" and have added the Maintenance
          Pack and are now unable to run "scoadmin printer" -> Printer ->
          Add Remote (ie. (UNIX) ) is not able to be highlighted to add new
          Remote Printers then the cause is /usr/lib/mkdev/rlp is missing.

          The file 'rlp' is kept in /usr/lib/lp/sysv/save and can be copied
          back to /usr/lib/mkdev/rlp

          This issue has been reported to SCO Engineering.

NOTE:
          If you need CUPS to be removed in order for you to use LPD as
          detailed above in the Maintenance Pack installation notes BUT also
          require SAMBA to be used then please note that the removal of CUPS
          will also remove a number of libraries vital to Samba from /usr/lib:

          libcups.a@
          libcups.so@
          libcups.so.2@
          libcupsimage.a@
          libcupsimage.so@
          libcupsimage.so.2@

          Which should be copied and placed back into /usr/lib after CUPS has
          been removed in order for SAMBA 3.0.20A to function otherwise you
          may see:

          # /etc/init.d/smb start
          dynamic linker : /usr/sbin/smbd : could not open libcups.so.2
     

DR/Migrate Sco SysV

1)      Reinstall SCO Base system

2)      Copy /etc/conf to a backup directory (ie: copy –ormv /etc/conf /usr/rcvr/conf)

3)      Remove all contents of /etc/conf (cd /etc/conf && rm –r /etc/conf/*)

4)      Change file creation mask on root filesystem (cd / && umask 0)

5)      Restore tape excluding /dev & /stand (cpio –icvdmufB< /dev/rStp0 “dev/*” “stand/*” )

6)      Copy ttytype & inittab to backup directory (ex: cp /etc/inittab /usr/rcvr/ )

7)      Use SCOadmin to relink the kernel

8)      If asked to rebuild environment answer (y)YES

9)      Put back inittab & ttytype form the backup (ex: cp /usr/rcvr/inittab  /etc/inittab)

     10)     Reboot & have a Beer

11

Totally forgot

Put the backup on a remote tape via ssh

#backup
   tar --verbose --exclude=/proc --exclude=/sys --exclude=/tmp --exclude=/mnt --totals -b2048 -jcpvf $SRC | ssh root@192.168.1.201 $(mt -f /dev/st0 rewind; cat > /dev/st0)

#restore
ssh root@192.168.1.201 "cat /dev/st0" | tar --exclude=/proc--exclude=/sys --exclude=/tmp --exclude=/mnt --totals -b2048 -jxpvf $DES