Here is a start:
Here is a start:
We have similar constraints inside our NethBOX (Nethesis appliances with NethServer). Our partition schema is:
- vda1: boot 500MB (kernel_limit should be set to 2)
- vda2: swap 1GB
- vda3: grow to end of disk
- About the RAM, I back @rgmhtt idea to go with ZRAM. Eventually, the user can switch to a SWAP file if needed. To start a 512MB zram should be fine.
- First partition should be a simple boot partition. We use XFS both for root and boot, but I think you can choose everything you want. My advice is to choose the same FS type for both partitions.
We do not handle UEFI machines, so we do not need an extra boot partition.
Just copy what CentOS or Fedora do
- Root partition at end is good
I’d like to publish a beta ARM image when 7.6 will be out!
Would a ZRAM of 512MB be acceptable with a 1GB ram arm board? Eventually I can test this on my Cubieboard2s.
The referenced howto is beyond my skills. I would need something much easier to set up. I am going to dig into what F29 is doing. I suspect they are running it as a systemd service.
As for Centos 7.6: Do you know of an anticipated availability date? Particularly for the arm image?
Oke, zram swap is the first option to pursuit. Do not know it and will investigate.
As said must get familiar with zram, but one of the motivating reasons to get rid of the swap partition is to be able to alter the swap size at will.
Oke we will keep the boot partition with a little bit of reluctance because still have found no arguments why it is preferred. Note most distributions do not use a boot partition in conjunction with u-boot.
In future it would be a preferred goal for me to boot all arm devices trough UEFI. This would make the boot process uniform for all architectures. Except (limited) UEFI functionality is provided by u-boot.
FC29 moved to this too:
I’ll guess you mean how fedora/cantos do it on x86-64. As said on arm it quite different and to my like.
That sounds like a plan, so we go for it!
If we are done before the centos-arm 7.6 nothing will stop us to release beta based on 7.5.1804 centos.
I will be investigating what F29 is doing. Perhaps we can follow that, though of course they will be on a newer version.
Is you mention, F29 makes a strong case for it for booting UEFI. I seem to remember a thread with Peter Robinson on this early in the F29 (then rawhide) release when I asked about the vfat partition.
Learn from F29-arm.
Push Centos-arm to adopt directly what makes sense. (see my zram post there).
Add whatever else we need think works here.
Or some such.
See my post on zram on the zram howto thread:
You must have some inside information on the progress of 7.6. For us regular Centos people all we hear is, “it will be released when it is ready.”
We don’t But we can manage the release of NethServer 7.6
I built a HD with:
Went through firstboot fine, I am logged in as root on the serial console on my Cubietruck. My IP address is 192.168.192.14.
From my F29 arm system on 192.168.192.12, I try to connect to:
And that does not work; I figured that I needed https. Wow is that default cert weird. I wonder if we can do better. If nothing else we need to move to ECDSA!
Set root’s password? What if I had already done this from the serial console, would I have to change it again?
I DO NOT APPRECIATE YOUR PASSWORD REQUIREMENTS. Fedora Initial Setup is very happy with my password of choice, giving it a good rating. And you don’t even give the password rules. Humph.
Set timezone? Why is not Detroit in the list here? It is in the tz file…
And I am in to the config…
I have two projects: Replace my ClearOS and my RSEL mailserver.
My Cubietruck should be up to replacing ClearOS for my 3 client systems. I will tackle that first as I decide which system to get for the mailserver replacement.
Get me what I need to convert to zram, and I will test switching to that as well…
Meanwhile I am going to figure out what I need to install to get the gnome DE and set up vncserver.
No, only if it is the default (Nethesis,1234)
Rules are quite strict, meanwhile i’m used to it…
Sorry, I do not know. have to look in to this.
A gross description:
- yum install zram
- disable swap (swapoff -a)
- edit /etc/fstab and comment out the “swap line”
- systemctl start zram-swap
- systemctl enable zram-swap
Thank you for testing and feedback!
I won’t get into password strict fights. I will see what I can do so it works for me…
timedatectl set-timezone America/Detroit
From the console works. Guess I just missed it in way your list works…
Other than it is ‘swapoff -a’, I am up now with zram. I will work out a sed command to comment out the swap line.
zramctl NAME ALGORITHM DISKSIZE DATA COMPR TOTAL STREAMS MOUNTPOINT /dev/zram0 lz4 959.4M 4K 64B 4K 2 <swap>
Nice that you already have zram in your repo! Send your rpm directly to Pablo so he can try it out.
Actually not hard:
sed -i -e "/swap/s/^/# /w /dev/stdout" /etc/fstab
With all this, you can wrap it up in a script and make it generally available as a Nethserver feature…
setting-up-a-pdc over in Support…
Do you have nethserver-dc available for the generic image?
seems to work for testing purpose for me.
Not sure to use it for production though…
I could use it for minor light non-critical tasks.
Thanks for testing!
I used a raspberry as fallback mirror webserver when the main webserver was down for maintenance. It’s enough to run a low perfomance home server.
I can also imagine using it as monitoring agent (with zabbix or nagios).
With better arm hardware than a raspberry there are no limits.
Monitoring agent on arm
I would definitly do that with a 4Go Ram Pi
I have several Raspbian Lite LAMP devices. I’m a bit oldschool when it comes to web servers. No virtual machines, no graphic interface, SSH only and no other service on the machine.
On day I’ll try with NS. But not on a small bug. Maybe on a supercharged Banana Pi?
I have had a Cubietruck running Redsleeve6 as my mailserver for 4 years. It processes ~5k messages per day and supports my very active access with 4 copies of Thunderbird via IMAP and one friend with POP.
The key here is that Cubies have native sata. The CT is 2GB memory duo core. It does get bogged down at times with the anti-virus scanning.
One of my goals is to replace this with a Nethserver mail server. I will probably also move to a quad core to better handle anti-virus scanning. Like the BananaPi M4 Ultra.
Warning if you want to run a desktop on Allwinner boards like Cubies and Bananas: they have the Mali GPU which is closed. None of the upstream kernels have support, so with GTK3, you end up in emulation mode. Fedora28-Xfce (GTK2) was decent on my Cubies. Fedora29-Xfce (GTK3) is bad. They are fine for servers, or if you use vncserver connection from your notebook (as I am doing). The Centos7-gnome works OK (probably GTK2).
I really like these small bugs!
A test I did this evening before starting over to get AD started right…
yum groupinstall “gnome desktop”
yum install vncserver
ssh in as user and run vncpasswd
configure vncserver for user
set firewall rule for ports 5901:5903
connect via vnc client
I am in! Also works for user root. Gives me a desktop to do more if I want. Logged out and stopped vncserver.
I would be nice if centos-arm had the Xfce desktop, but not all the modules are there for armhfp.
So that is enough for tonight. Shutdown and start over and this time get AD right…
Or righter. Ready, fire, aim, aim, aim…