I was bored and I installed Slackware just for fun (could not wait for 15) and because somebody “complained” nobody on this board was using Slackware. Now @Ulfnic can add it to DLN’s distro database.
Look at the package count! It is slim but do not forget Slackware comes with everything so I had to trim it down and I only wanted Xfce without old KDE 4. That was taking the time of an Arch install because otherwise Slackware is not really that difficult if you are in love with an ncurses installer from the 90s and using cfdisk to partition your drive. It is truly KISS and I love it, I mean I am that old school anyway.
This thing and old Xfce 4.12 flies! It takes 200 MB of RAM on boot!!! That is how I love my Linux, truly lightweight and I have all the apps I need. I added one AppImage for Standard Notes otherwise I still did not compile* any package but everything is installed if you need to compile from source.
Of course the install would be faster if I were to take the recommended way of installing everything but who wants that, no fun to me and I only want what I really need.
I should probably write a how-to.
*Actually not true. I used Slackbuilds for some stuff but that does not count to me as truly compiling from source because it uses scripts to compile the software but it is a great source for software that is not in the official Slackware repositories and the packages are usually of good quality.
Finally got around to rebuilding my home lab and replacing CentOS 7 with Debian 10.
Got KVM installed, configured a bridge for the VMs. Created a base image and now have 4 VM’s running with Ansible and a Unifi network controller on one node. Grafana and Influxdb running on another node (with all nodes running Telegraf). One node will be hosting OpenVAS, and the last node will be a Python dev server.
The installation and configuration of Grafana, Influxdb, and Telegraf (on all 4 VMs and the VM host) took less than 20 minutes. Although, this is not my first time working with these monitoring tools.
Now you have me curious… is the transition to Debian because it’ll serve you better than CentOS or is it because you consider it a better option to going Rocky/Alma? I’m really spunging up these opinions at the moment.
My concern is having my operating system and it’s updates directly attached to my personal information. It’s not a policy that’s mindful of the kind of things Snowden talks about. For example it’s easy for a gov’t to get away with compelling a company or employee to poison an update going to one person, hard if it has to sit in a public repo. It also opens the door to Red Hat being pressured to cancel accounts of servers that host unpopular speech (or simply belonging to a company where the CEO said something bad 10yrs ago) as it’s becoming a trendy thing to push companies to do.
While i’m in the lucky category of being very boring to everyone, it’s not a precedent i’d prefer to reward and I think mandated accounts are best left for Windows 11.
That said… the reason Red Hat set it up this way is understandable. Red Hat is generally pure of spirit and if I had to trust a company Red Hat is a very easy choice (my laptop runs on CentOS) but a little bell in my head just won’t let me attach my ability to use software to a central authority.
For a company however, it can make a lot of sense and the support is a massive plus.
That’s a big question that I spent a lot of time considering.
I originally choose CentOS because I work with RHEL at work. CentOS/RHEL is a great server platform, but there are others that are equally good.
My first thought was to move to Rocky, but it wasn’t ready at the time so I waited and thought about what Redhat had just done to the community. In the end, I choose Debian because it isn’t owned directly by any company. My intention was to move to a distro that had the least amount of commercial influence. This is what led me to Debian.
Keep in mind that my move to Debian is only for the server side. I am still VERY happy with Solus Linux for all of my desktop needs.
These are my personal thoughts and opinion. I respect that others have different thoughts on the matter.
I do accept that Redhat does a lot for the community. For example, Redhat has done a lot to further KVM. If my memory serves me correctly, libvirt came from Redhat. So, I am thankful for that.
Firstly the registration for a subscription. Then I have to renew the subscription every year. It seems overly tedious and complicated to me and of course they want my data.
Second, Red Hat still can change its policy whenever they see fit. I have concerns with that.
Everything above is not a problem per se but not what I want on my personal machine. For enterprise
I used my Ryzen 3 2200G now for 2.3 years on its default settings. It worked fine, but the boot times for my Linux VMs slowly increased from say ~10 seconds to ~11 seconds and reboot times were ~9.5 seconds. I use OpenZFS on Ubuntu 21.04. Since 98% of all drive IO during reboot is done from the memory cache (L1ARC) instead of the nvme-SSD. I realized, that a relative large part of the boot time is caused by the CPU instead of my nvme SSD (3400MB/s).
Today I manually clocked my CPU to its boost frequency of 3.7GHz, with temps in the 49-65 range, so it looks fine, the fan speed is 200 rpm higher. Ubuntu Mate 21.10 booted in 8.5 seconds faster than ever and the reboot time from L1ARC is 7.3 seconds.
I think I gained the reaction times (2msec) the older Ryzen CPUs need, when boosting its frequency a few times during that boot process and of course the CPU now runs constantly at its boost frequency. .
I have been running Zorin 16 for a week or so on my very slow Pentium branded CPU on my second laptop with 4GB Ram. So far I like it a lot, Previously I ran Elementary 6 for a week which was too slow (Flatpacks needing fast cpu cores maybe). Before that I ran Ubuntu Mate which I loved. Maybe I’ll go back to UM but want to give Zorin a bit longer before deciding.
This weekend i saved a lenovo thinkpad from being forgotten for ever. It was sitting on a desk, disregarded for quite some time now. I took it with me, dismantled and cleaned it, and now i’ve ordered a brandnew 2,5" hd, which will receive a distro upon arrival. I just don’t know which one yet.
I killed a Dell Latitude E5540 laptop installing Debian Sid a few weeks back, that was traumitizing, i instantly fell in love with KDE and violently had it ripped out my hands.
Eventually we realized that we had an extended warranty. I returned the expired Dell for a full refund, and got an HP Elite Book 830. I had 3 distros installed on the Dell on a 1Tb SSD, and putting in the HP, after some patience, i had all three distros working again! Sid would not initiate the wifi. I ended up struggling with an Arch install to replace it. Grub didnt intall right, but grub rescue saved me!! I installed KDE, am thoroughly enjoy it. Being an XFCE user for years, moving to KDE was un-natural, but its happening…
Accidentally ended up on PopOS for a spare laptop we had laying around, which will become the general “I need a real web browser” computer for the family.
Realtek Wi-Fi cards suck….
Tried to install Manjaro, but I couldn’t get the Wi-Fi to connect, so I tried Fedora and that worked for a bit until the laptop slept, I eventually remembered that Pop was a thing and might be worth trying. Wi-Fi card is still kinda stupid, and takes forever to reconnect if it sleeps or goes into airplane mode, but it does reconnect faster than Fedora or Manjaro.
It looks good too, and seems to be pretty solid overall, lots of things to like about it, especially for people that don’t want to mess with stuff and just need a web browser.
I don’t think it’s something I could use on a daily basis, or as a daily driver, (as a member of the KDE Kmafia) but it’s pretty darn good.
I did some stuff, that I postponed for some time, like:
I did subscribe to additional free ESM updates for Ubuntu 14.04 LTS and 16.04 LTS and I did upgrade those Virtual Machines (VM).
I did rearrange the proposed pre-released development editions of Ubuntu 21.10. I have 4 Development Editions in a VM. I switched them all back to normal development editions. I had some crashes and after seeing that the crash dumps had been uploaded, I switched back by rolling the stuff back to 2 weeks ago (advantage of the Host running OpenZFS), change the update settings and re-updated the 2 distros.
I upgraded the VBox Windows 10 Pro VM to Windows 11 Pro. Note that my Virtualbox Host OS is Ubuntu 21.04 running on OpenZFS 2.0.
My Ryzen 3 2200G is NOT supported by Windows 11, so I can’t use Setting → Update. There Microsoft tells me that my PC/VM is not compatible with Windows 11. For that upgrade I did the following:
I cloned the VBox Windows 10 Pro VM, changing name and location already. Note that my Windows 10 Pro is already since 2018 part of the Windows Insider Program;
I downloaded the official ISO from the Microsoft web site.
I mounted the ISO in the Windows 10 Clone VM.
I did run “setup” from the 4.5GB ISO file and the system completely upgraded in say an hour.
TPM 2.0. fTPM 2.0 is still switched off in the BIOS of the Host;
The original Windows 10 Pro VM did not use UEFI boot;
The disk has the good old MBR lay-out, see picture above
Note that the 14nm Ryzen 3 2200G is officially NOT supported by Windows 11.
I did run settings → update in Windows 11 and that seemed to work. There were no updates, probably because the updates were downloaded during the installation. For the moment I feel happy, because my collection of Windows VMs from 1.04 to 11 is complete again
All existing Windows 10 Apps did work and that included my Virtualbox Guest Additions 6.1.26. Both 3D acceleration for the video and the shared folders worked. I will keep you informed by adding stuff using “reply”. I still have to check that I really get updates and I want to have Windows 11 activated. On Saturday I backup everything to my laptop. I will try to use the laptop Windows 7 key to activate Windows 11 on that laptop, interesting since the CPU is an i5-2520M.
A busy day Friday, I have run the updates for 2/3 of my Virtual Machines (VM). It takes 5 to 6 hours. My Ryzen3 22200G is really busy, when I update 2 Windows VMs at the same time, For an hour the system runs a ~100% CPU load. I have recently overclocked the CPU to 3.7GHz, that helps somewhat.
The photo shows the update of Windows 10 and you can see in Conky on the right that almost all 16GB of memory is used and CPU loads are close to 100%. The CPU temperature is in the 70°C below the 80°C, at which the CPU starts throttling. To proof that I run two Windows Update see the photo with Windows 8.1 below. Win 8.1 is ready with the limited updates and it now is cleaning the disk and especially deleting the system files and restore points. Note that both Windows VMs take approx 50% of the CPU time, see in the right upper corner of the Windows window.
After a while the updates get boring, so I decided to play some music. The Host OS is a minimal Install of Ubuntu 21.04, so I need another VM to play music. So I started Windows XP Home.
See it playing my music. The music is playing without any stutters. I’m impressed by the Linux scheduler I run the VMs with 4 cores each, except Win XP Home, that only supports one CPU. Note that the 16GB is now completely used for VRAM 1GB; ZFS Cache 4GB; Windows 10 3GB; Windows 8.1 3GB and Windows XP 1GB. The remainder is used by the Host OS; Virtualbox and ZFS.
All 3 VMs are run from the dpool datapool, which consists of 2 striped 500GB HDD partitions (Raid-0) supported by a 95GB SSD cache (L2ARC and ZIL). The Windows 11 VM and the other Windows 10 VM from my previous entry here run from my nvme-SSD.
The VMs react slow from time to time running 2 or more updates at the same time, so I consider a Ryzen 5 4600G or 5600G, next year when the prices are right for me, due to the introduction of ZEN 4.
I know it’s a joke but I had a similar problem recently. Days later I remembered iostat and found the culprit process baloo churning up 100s of GB of my disk space.
