Tag Archives: linux

Podman – unable to pull image

Today, while pulling the latest Nextcloud container image from the docker.io repository I noticed that it is failing. Looking at the read: connection reset by peer error I assume it could be primarily due to network failure; maybe poor quality of the connection.

Note that Mauritius is far from "everywhere" else and connectivity issues are nothing new.

ish@coffee-bar:~> podman pull docker.io/library/nextcloud
Trying to pull docker.io/library/nextcloud...
Getting image source signatures
Copying blob e0276193a084 done  
Copying blob eb2d00c10344 done  
Copying blob 3a60f364b0c5 done  
Copying blob e0d3d1244592 done  
Copying blob f54006e0dc29 done  
Copying blob 8559a31e96f4 done  
Copying blob b22875b95a2a done  
Copying blob f65316e96b10 done  
Copying blob 3e309988c00b done  
Copying blob 0c78caf16ec3 [======================================] 13.1MiB / 13.2MiB
Copying blob 4fc30aae7ee5 done  
Copying blob 37b016cacdc6 done  
Copying blob fd56bf3cc539 done  
Copying blob e3cd35f544b5 done  
Copying blob 467fea8f6f80 done  
Copying blob 0dc7444f9282 done  
Copying blob 547ae6684264 done  
Copying blob 3139b6de5be7 done  
Copying blob 00e77223b529 done  
Copying blob 2b37e3b3a856 done  
  read tcp 192.168.100.6:33536->104.18.124.25:443: read: connection reset by peer
Error: error pulling image "docker.io/library/nextcloud": unable to pull docker.io/library/nextcloud: unable to pull image: Error writing blob: error storing blob to file "/var/tmp/storage720474498/10": read tcp 192.168.100.6:33536->104.18.124.25:443: read: connection reset by peer

Podman does not retry to copy the image in case of failures.

A few days ago there was a suggestion to implement a similar feature in Podman that is present in Buildah, which provides the image copy retry functionality.

For the curious, the implementation can be seen here.

Container images that are pulled by Buildah are stored in the local repository which can also by accessed by Podman, so that's an advantage. I tried pulling the Nextcloud container image using Buildah and it completed successfully.

ish@coffee-bar:~> buildah pull nextcloud
Getting image source signatures
Copying blob e0d3d1244592 done  
Copying blob 8559a31e96f4 done  
Copying blob eb2d00c10344 done  
Copying blob 3a60f364b0c5 done  
Copying blob f54006e0dc29 done  
Copying blob e0276193a084 done  
Copying blob f65316e96b10 done  
Copying blob b22875b95a2a done  
Copying blob 3e309988c00b done  
Copying blob 0c78caf16ec3 done  
Copying blob 4fc30aae7ee5 done  
Copying blob 37b016cacdc6 done  
Copying blob e3cd35f544b5 done  
Copying blob fd56bf3cc539 done  
Copying blob 467fea8f6f80 done  
Copying blob 0dc7444f9282 done  
Copying blob 547ae6684264 done  
Copying blob 2b37e3b3a856 done  
Copying blob 00e77223b529 done  
Copying blob 3139b6de5be7 done  
Copying config 327476ebe3 done  
Writing manifest to image destination
Storing signatures
327476ebe3280c7b570d8463edd136956eab120959976b643cb7dbfaa73f98c1

Now, the downloaded container image is also accessible by Podman.

ish@coffee-bar:~> podman images
REPOSITORY                     TAG      IMAGE ID       CREATED        SIZE
docker.io/libreoffice/online   latest   0586fecfa3c1   28 hours ago   2.84 GB
docker.io/library/nextcloud    latest   327476ebe328   3 days ago     774 MB

To conclude, while we are waiting that a retryCopyImage function is available in Podman, we can use Buildah to pull container images that are troublesome due to network issues.

A quick intro to Podman by Estu Fardani

Estu Fardani is a helpful & jovial fellow whom I met at the openSUSE Asia Summit last year in Bali, Indonesia. Recently, for the openSUSE Virtual Summit, Estu did a short presentation on Podman. His presentation video is available on YouTube and slides deck available on Google Slides.

While I am not a fan of alias docker=podman I believe Estu added that part in his slide below to make the Podman transition a bit smoother for users already familiar with Docker commands.

Slide from Estu's presentation

Note that the latest release of openSUSE Leap, i.e version 15.2, comes with support for Podman through the libcontainers-common package.

Previously, if one needed to experiment with Podman and/or deploy in production then openSUSE Tumbleweed, Kubic or MicroOS were the supported choices. Now, one may deploy Podman containers on their existing Leap infrastructure (after upgrading to the latest version 15.2).

कुछ बातें openSUSE Leap 15.2 के बारे में जानें

२ जुलाई २०२० को openSUSE Leap 15.2 संस्करण उपलब्ध हुआ।

openSUSE Leap एक निशुल्क और लिनक्स-आधारित ऑपरेटिंग सिस्टम है जो आप अपने पीसी, लैपटॉप या सर्वर पर इस्तेमाल कर सकते है। Leap 15.2 और SUSE Linux Enterprise एक समान कोडबेस उपयोग करते है जिस से openSUSE Leap 15.2 की स्थिरता और ज्यादा मज़बूत होती है।

यह विशेष संस्करण कृत्रिम बुद्धिमत्ता, यानी आर्टिफिशियल इंटेलिजेंस, को अधिक आसानी से सुलभ बनाता है। निम्नलिखित पैकेज ऑफिशल रिपॉजिटरीज से प्राप्त किए जा सकते हैं।

इस संस्करण में अधिक कंटेनर टूलस् शामिल है, जैसे कि कुबनेटिस् (जो पहली बार के लिए Leap संस्करण में उपलब्ध हुआ है) हेल्म और सिल्यम

Convert .VOB files to .MP4

I had a bunch of video files in the .VOB format which I copied from a few DVDs a couple of years back. Uploading them on a cloud storage service (e.g Google Drive, OneDrive etc) for safe keeping is nice but not practical since they cannot be played directly through the cloud service's video player. The latter would play .MP4 video files.

FFMPEG comes handy here to convert the video files.

ffmpeg -i bla.vob -b:v 3000k -b:a 256k bla.mp4

I used mediainfo to find the maximum video & audio bit rate.

mediainfo bla.vob

Running Puddletag on Linux distributions with limited Python 2 support

I am a big fan of Puddletag, an audio tag editor for Linux similar to the Mp3tag Windows program. As seen above, Puddletag uses a spreadsheet-like layout which makes it quite unique in the Linux world and matches 100% with how I like tagging to be. Puddletag and I are on the same wavelength and I have been using it for some years now to curate my extensive music collection.

Unfortunately, Puddletag was written in Python 2 and relies on a number of Python 2 libraries. When I upgraded to Fedora Linux 32 a few days ago, I lost a number of those libraries as well as Puddletag (as Python 2 reached its end of life (EOL) on 1st January 2020).

Fortunately, the Linux community came to the rescue and created a corresponding AppImage which works on all new Linux distributions like Fedora 32 or Ubuntu 20.04. Get the AppImage here courtesy of Patsim and have fun!

I’ve tested it in Fedora Linux 32 and it works great!

Libreoffice with Flatpak: Adding dictionaries for other languages

I generally write in either English, French or Mauritian Kreol.

I also use Libreoffice when I need a word processor or a spreadsheet (or, even, sometimes, a drawing software). Lately, I have discovered Flatpak (“The Future of Apps on Linux”) and Flathub (“An App Store for Linux”) and I am sold. Installing the latest version of Libreoffice from Flathub using Flatpak is a simple:

$ flatpak install flathub org.libreoffice.LibreOffice

I have noticed though that this command tends to only install the English dictionaries for spellchecking. I didn’t know how to install more dictionaries to Libreoffice so I asked on the official Flathub forum. And, fortunately, someone from the community pointed me towards the solution. The idea is to get some information about the Libreoffice installation:

$ flatpak info org.libreoffice.LibreOffice

LibreOffice - The LibreOffice productivity suite

ID: org.libreoffice.LibreOffice
Ref: app/org.libreoffice.LibreOffice/x86_64/stable
Arch: x86_64
Branch: stable
Version: 6.4.3.2
License: MPL-2.0
Origin: flathub
Collection: org.flathub.Stable
Installation: system
Installed: 686.0MB
Runtime: org.freedesktop.Platform/x86_64/19.08
Sdk: org.freedesktop.Sdk/x86_64/19.08

Commit: ddcb114395acb30f633a06fd065598ace0fbe4330c49a784a50911b0222f5f95
Parent: fa891c405f685f7859e2bb623b29db5cdb3e9e1d80d8c31f30a5d21edcc9a3eb
Subject: Update to libreoffice-6.4.3.2 (5a34256e) Date: 2020-04-16 18:32:28 +0000

Notice that the runtime is org.freedesktop.Platform/x86_64/19.08 which implies that org.freedesktop.Platform.Locale/x86_64/19.08 contains all the dictionaries. But, as pointed out by stbergmann in the forum,

The *.Locale extensions are special, in that flatpak by default only downloads that part of such an extension that matches the users current system locale, while the –reinstall hack unconditionally downloads all of it. There obviously needs to be usability improvements here, as this appears to be a common issue for users.

And this is the –reinstall hack he is talking about:

$ flatpak install --reinstall flathub org.freedesktop.Platform.Locale/x86_64/19.08

and, after a few seconds, the dictionary issue is cured. Now, in addition to English, you have French plus all the other languages. Enjoy :-)

Managing the Linux /home directory is about to change

systemd 245 came out on the 6th of March 2020.

In this release, among many changes brought, a notable one is that regarding how we see and think of the Linux /home directory. A new service systemd-homed.service has been added, whose role is to manage home directories.

Traditionally, user information on Linux systems have recorded in the /etc/passwd and /etc/shadow files. The former contains details about the user name, id, group id, shell, home directory, among other details, while the latter contains the user password information.

On the other hand systemd-homed.service stores user information in a ~/.identity file in JSON format. The choice of JSON was mainly due to its popularity and easiness to process with the majority of programming languages.

Thus, every directory managed by systemd-homed.service contains both the user information and the user data.

To create, remove or change home directories one would use the homectl command.

$ sudo homectl create john --real-name="John Doe" -G wheel --disk-size=500M

What if you could carry your home directory in a pendrive and work from any (Linux) machine as if you are "home" ?

$ sudo homectl create john --real-name="John Doe" --image-path=/dev/disk/by-id/usb-SanDisk_Ultra_Fit_476fff954b2b5c44-0:0 --tasks-max=500

The above command creates the user john in a pendrive and assigns a maximum of 500 concurrent task to him.

systemd-homed.service also allows you to have an encrypted home directory, but that, I will write about on another day. 😉


Twitter/Facebook card image source: auroria.io

Disable tracker in Gnome if you do not need it

According to its website, “Tracker is a filesystem indexer, metadata storage system and search tool.”

“By using Tracker, you no longer have to remember where you’ve left your files. To locate a file you only need to remember something about it, such as a word in the document or the artist of the song. This is because as well as searching for files in the traditional way, by name and location, Tracker searches files’ contents and metadata.”

That’s great I guess for some people.

For me, Tracker is a massive pain as I have a lot of files, some of them huge and a lot of them being binary files (think images and photos, music files, videos, LibreOffice files, PDF documents, etc.) On my Fedora Linux installation, which is now running in a virtual machine on a brand new MacBook Pro, the Tracker extractor and indexer uses a lot of CPU resources. Too much in fact for something that I never ever use.

You see, I am an “informatician” (i.e. a computer scientist — a very bad designation — think calling a surgeon a scalpel scientist) and, as such, I make it a must to properly save all my files in folders and subfolders. I also name the files very consistently with dates, sensible use of words, etc. for proper sorting and, hence, searching.

So I don’t need Tracker as I pretty much can narrow down to the single file I need in a few seconds.

How to disable Tracker in Gnome

I found the following set of commands online which disable Tracker:

    gsettings set org.freedesktop.Tracker.Miner.Files enable-monitors false
    gsettings set org.freedesktop.Tracker.Miner.Files ignored-files "['*']"
    gsettings set org.freedesktop.Tracker.Miner.Files crawling-interval -2

Looking at the source code of Tracker,  the -2 indicates that Tracker should be disabled. The next step is to kill all current Tracker processes:

    pkill tracker

and, finally, clear the Tracker cache

    rm -rf ~/.cache/tracker

Playing Quake again after 23 years

Quake was born when Christina and I were studying in France in 1996.

Like most people we knew, we fell in love with the game: the graphics were amazing (and we even bought a 3dfx Voodoo 2 to run it), the audio was great and the atmosphere phenomenal. As a matter of fact, Quake is one of the few games I completed. I ran it on Windows 95.

Fast forward 23 years

Today, my laptop runs Fedora Linux 29 and I only use its integrated Intel graphics chipset (despite the laptop having an nVidia chipset as well). This morning, I stumbled upon Quaddicted:

From there, I quickly discovered that, since Quake was released as open source software in 2012, a number of people have embarked on correcting bugs and enhancing the Quake engine. According to Quaddicted (and who am I to disagree!), the best engine for Linux is Quakespasm.

I quickly downloaded it and found an executable called quakespasm-sdl2 which I quickly ran. I got an error message telling me that a folder id1 was not found.

I remembered having zipped my Quake installation years and years ago and, because I am a very methodical guy, took only 5 seconds to find the ZIP. It had the id1 folder in it which I quickly copied to the Quakespasm folder. I got the Quake music. I ran quakespasm-sdl2 again and, lo and behold, Quake started in all its glory. I quickly selected the full screen option and opened a save game from decades ago and it worked! This is what you can see in the top screenshot.

But quake is more than quake…

While reading Quaddicted, a stumbled upon Quake Injector, a tool written in Java which allows you to easily download and run 3rd party maps in Quake. Think maps as levels, distinct from the Quake default levels, but as great and sometimes greater to play.

[Caveat: Quake Injector did not run at first as I had Java 11 on my Linux laptop. I had to switch to Java 8 (which I also had but was inactive)]

As soon as Quake Injector started, I installed the very highly rated DaMaul6 map and this is where I am now. Wish me luck as DaMaul6 is known to be very difficult.

A journey from Unison to Syncthing and back to Unison

I generally work on my laptop, an old Dell Inspiron 15 still running Fedora Linux 29 quite well. At home, I also tend to work on an Apple iMac (still) running macOS El Capitan (which I intend to upgrade to High Sierra soon — unfortunately, Apple has decided that this iMac cannot run Mojave).

Since the very beginning, I needed to have a proper file synchronisation process for these two computers and cloud solutions such as Google Drive or Dropbox were not options as I had many many Gigabytes to synchronise and this would have been very expensive to be done online.

For 1-2 years, I used Unison, a file synchronisation utility written by one of my Computer Science heroes, Benjamin Pierce, author of the quite renowned book “Types and Programming Languages” which, I have to confess, I have not yet read because, well, I never had the chance to stumble upon it. Anyway, Unison is a tool which does two-way synchronisation between computers and every time there is an inconsistency (a new file, a deleted file, a change, etc.) ask the user to manually choose on the action to make. One important of Unison is that one has to run it manually i.e. the synchronisation only happens when initiated by the user. For example, I like to run Unison every afternoon when I get back home after one day of work.

One year ago, I stopped using Unison. The reason was that I wanted to have real time synchronisation between the two computers i.e. as soon as one changes, the other one is updated (if on of course.) I settled on Syncthing. I tried two different ways of using Syncthing (with a central repository or in peer-to-peer mode) and both worked as expected. But after one year, I have to say that I have two issues with Syncthing. First of all, changes are detected through the inotify mechanism provided by Linux and the macOS kernel. As soon as a file is changed, Syncthing knows about it and can propagate the change to the other computer. All good in principle. But this also means that all errors (a file deleted by mistake, a file overwritten by mistake, etc.) are also immediately propagated. So I used different kinds of versioning techniques to make sure that backups are kept in these situations. But, nevertheless, I slowly realised that real-time propagation might not be a good idea. The second issue is that Syncthing uses a lot of CPU.

Two days ago, I stopped the Syncthing service on all my computers and revived my Unison installation. And, believe it or not, Unison works great for my use case: I can synchronise my computers when I want (after work for instance) and I have the possibility (if there are incoherences) to think deep about which version to keep. For me, and because my files are very valuable to me, this is a much better process.

[At this point, let me point out that I use Unison for synchronisation and I use regular rsync for backup on other devices. It is important to understand the difference between these two processes. Synchronisation is not always needed but backups definitely. Especially when it is compliant with the 3-2-1 strategy.]

What about you? What do you use for synchronising multiple computers when cloud synchronisation is not possible? Do you use Unison? Syncthing? Something else? Why?