Definite's Extractor

My findings on Life, Linux, Open Source, and so on.

Category Archives: *nix

HOWTO: Configure Time (NTP) for Windows Server 2016 Using Command Line (cmd)

This is aimed to be complete steps for configuring the time service (NTP) for Windows Sever 2016. This is especially useful should you want to setup your own Active Directory Domain Controller (ADDC).

0. Networking Check

Ensure you can reach the NTP server. Assuming you are using :


1. Set the Time Zone

By default, windows just assume a time zone and show you the system clock. However, the default time zone is not necessary correct.

1.1. Check the Current Timezone

tzutil /g

1.2 (Optional) Query Available Time Zone

tzutil /l

1.3 (Optional) Set Time Zone

tzutil /s "E. Australia Standard Time"

2. Set the w32time service

2.1 Configure NTP Service

w32tm /config /manualpeerlist:"" /syncfromflags:manual /reliable:no /update

2.2 Confirm that NTP Configuration

List the peers

w32tm /query /peers
The result looks like:
#Peers 3
State: Active
Time Remaining: 123.4567890s
Mode: 1 (Symmetric Active)

List the time source

w32tm /query /source
The result looks like:

2.3 Resync

w32tm /resync
The result looks like:
Sending resync command to local computer
The command completed successfully.

2.4 Indicate that NTP service is now reliable

w32tm /config /reliable:yes /update

2.5 Confirm status

w32tm /query /status
The result should looks like:
Leap Indicator: 0(no warning)

This should be it.

Build in epel8 branch with fedpkg

This article assumes you have commit right to the package, but you don’t have the epel8 branch in

  1. PKG=<YourPackage>
  2. fedpkg clone $PKG # if you have not done this
  3. cd $PKG # if you have not done this
  4. git branch epel8
  5. # Make the branch work with epel8
  6. git commit
  7. git push –set-upstream origin epel8
  8. fedpkg build
  9. Go to bodhi and create -> New update


The “Proper” Way to Use Microsoft Word Master Document

Long story short, don’t use it, use field IncludeText instead.

As a DevOps/software engineer, I prefer to keep my files (even the MS word documents) consistent and synchronised.

I am looking for jobs and sending my resume to two kinds of organisations: public sector and industry. Public sector jobs usually require referees, while in industry, referees are not mandatory. So I have two types of resumes: with or without referees.

Why don’t I just use the one with referees? Some international recruiters are not aware of time differences and just call you in the most inconvenient time. I have received a call at 4:30 am, when all of my family are sleeping. So that’s why I am reluctant to include referees in my industry resume.

To keep both branches in sync, I though the master document was a good idea, until I saved and opened the master document, and I see mess, just as describe in Why Master Documents corrupt. And I saw an interesting quote:

There are two kinds of Master Documents: Those that are corrupt and those that will be corrupt soon. – John McGhie

I have also tried auto text in “building block”. The main problem of that is: I have hard time to locate them. I cannot remember whether I save it in or . When I need to update contacts several months later, I will probably forget they are in auto text.

Eventually, I found that the field IncludeText actually helps. But there are CATCHES:

  1. The insert field command is also not easy to find in UI. Search “Insert field” is more reliable. (The search field is at “Tell me what you want to do”)
  2. You need to manually input the filename, there are NO file dialog for you to choose.
  3. If the document is on OneDrive, the full path in local drive WON’T WORK. You need related path to OneDrive\Documents. For example. If the sub document, referees.docx is in the directory OneDrive\Job and your main document is also at OneDrive\Job, you need to insert ../Job/referees.doc to the field IncludeText in your main document.
  4. URLs in sub documents need to be well-formatted. In other words, for email addresses, WON’T work, you need
  5. Even if you do enter the correct URLs, MS Word like to show them as it pleases. It is up to MS Word to decide whether to show the URL as URL or plain text.

I uses Office 365, so version-wise it supposes to be always the latest. Perhaps I should have written the sub-documents as HTML+CSS and version control them.

HOWTO: sign dkms kernel module for UEFI secure boot

If you are stuck with UEFI secure boot, and desperately need to build and install a custom kernel module, then you need to know how to sign your kernel module.

To sign a dkms custom kernel, please refer my Git repository: dkms-module-util

Fix NTFS “unsupported reparse point”

Reparse point is an application-defined data that associate with a file or directory.  So it can be anything from symbolic link to mount point, and many more.

A comment in ntfs: unsupported reparse point says that plugins from Advanced NTFS-3G Features fix the problem. My own experience: works, but binary provided in is not.

HOWTO: Unwatch packages in Fedora package source (

If you don’t want to get messages from certain Fedora packages:

  1. Login  to
  2. Click Watchlist
  3. Click the package you want to unwatch, this will bring you to package Overview page
  4. There is an Unwatch  pull down which looks like the below picture. Click it and select UnwatchPackage Overview page

Note that you CANNOT unwatch the packages you maintained.

Legacy Python str.format() gotcha

In Python 2.7, str.format() works with empty string like:

python -c "print '{}'.format('')"

However, in Python < 2.6,  you will see:

ValueError: zero length field name in format

Workaround? Use %-formatting like:

python -c "print '%s' % ''"


Indent in Fedora-Flavored Markdown

I cannot find any documents about the indent in Fedora-Flavored Markdown. Yet I figure out following:

You need 4 spaces to indent. The rest is same with this Markdown Cheatsheet.

For example

- Upstream update to 1.10.0
    + Bug fixes:
        - Log to file defaults to unwritable location
        - Losing GUI config when restarting the OS

    + Enhancements:
        - CLI argument to control screen lock feature
        - Customizable size limit on clipboard sharing

- BuildRequires qt5-qtbase-devel


  • Upstream update to 1.10.0
    • Bug fixes:
      • Log to file defaults to unwritable location
      • Losing GUI config when restarting the OS
    • Enhancements:
      • CLI argument to control screen lock feature
      • Customizable size limit on clipboard sharing
  • BuildRequires qt5-qtbase-devel

See for the actual result.

Debugging Customizable Kickstart File

Kickstart file automates the installation Linux system. It does not provide a convenient way to pass parameters. Luckily,  passing kernel parameters to %pre installation section works. This method requires you to write your kickstart generation script in %pre section, then in the command section, you call the generated kickstart file.

Unfortunately, this mean it is more likely to fail if the parameters are not correct. How to debug? If you can at lease reach the installer, this document may point you direction.

1. Get the shell access

After anaconda the installer starts,  there are several console tabs, the ones that are useful are:

  • main: the main information screen (Alt-F1 or Ctrl-Alt-F1)
  • Shell: the shell access (Alt-F2 or Ctrl-Alt-F2)
  • GUI: show the install progress if graphical mode is not disabled (Alt-F6 or Ctrl-Alt-F6)

For now, shell access is the focus. Switch to Shell tab by pressing Alt-F2 or Crt-Alt-F2.

2. Inspect parameters by looking at /proc/cmdfile

Just cat /proc/cmdfile to inspect whether the parameters looks right. Personally, I do not want to write complex parser, thus I use quote_plus() from python to encode tricky characters like space, line feed, $ and so on.

3. Inspect the generated scripts and log files

The first thing to check is generated kickstart file. Then check following interesting files in /tmp

  • ks-script-XXX: The script in  section %pre  will be here. Inspect it, and even run it to see how it works or not
  • *.log: various of log files

Enjoy the debugging. 🙂

Solved: nginx module incompatible

I encountered when I run nginx -t:

nginx: [emerg] module
version x instead of y in 

It indicated that the module version is incompatible.
If nginx was installed using yum or dnf, use following command to set the version:

yum downgrade nginx{,-all-modules,-filesystem,-mod-*}-[version]-[release]