New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Los Angeles Double Dedicated Debian / FreeBSD Shell Accounts from MetalVPS
MetalVPS got a couple of E3s in Los Angeles from OneProvider.
E3-1230 v2
16 GB DDR3 (Not ECC)
1 x 1 TB HDD
1 Gbps unlimited
1 IPv4
No IPv6
Debian 12.4 / FreeBSD 14.0
Set to cancel in a month, unless cancellation is revoked.
Anyone want a Debian or a FreeBSD shell account, maybe with sudo? 🍔
You have to pay something. The servers cost $17.99/month each.
You need to tell @Not_Oles who you are.
No service level promises. Not for business use. Just for learning and fun.
Please get in touch via the email on @Not_Oles profile, posting here, or PM.
Probably this is way too crazy!
Best wishes!
Tom
Thanked by 1noob404
Comments
Update February 11, 2024!
Right now these two Los Angeles servers are very happily running FreeBSD 14.0! Nevertheless, it would be easy to return one or both servers to Debian, or change to Ubuntu, Alma, or Rocky, or something else.
Both these LA servers are set to cancel on February 26, but that's still 15 days away. Nobody else but me is using them. So if you want a peaceful place to try something, with a decent, dedicated processor, and a pretty good network, now might be a great time to come aboard.
Thanks to help from @Crab, I had tons of fun learning a little about installing and updating FreeBSD on these servers! I feel like I already got way more than full value out of the funds I invested! Therefore, maybe I could give you an account for super cheap via Paypal or Zelle. How much do you want to pay?
In addition to these LA dedicated servers, MetalVPS has more dedicated servers in Washington state and in Germany. If you want accounts in additional locations, we can set them up easily!
Friendly Low End greetings!
Tom
I'll pay $0.2 a month for a shell.
I'll pay the cost for one server. I'm ready to come aboard if you'll allow it, Skipper (Professor, Millionaire?), as long you don't hit me with your hat if I crash it into the island.
Have you thought about what you could launch from the shell? Is there anything that the server could do which could not be launched from the shell?
All right, you are in. If you behave, I will do my best about my hat.
Please send me half a month for one server, ~ $9.00 via Paypal or Zelle. Paypal can be via the email address on my LET profile. If you want to use Zelle, let me know, and I will give you payment data.
Also, if you do not mind, please send me or post your ssh ed25519 public key.
Last, but not least, do you want FreeBSD? Or Debian, or something else?
Welcome aboard!
ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIN4emHkCbrDflFRNfNMMVKP9yjxxT2NdqPGKAiEc3AN7 michael@optiplex790-2023-03-09
Debian 12 please. If you want to add a hostname, you can use privacynonprofit.org or I can set it up later.
I don't mind sharing as long as all the aforementioned terms are followed AND I get to learn from whatever you do.
Thanks for the opportunity, @Not_Oles! Looking forward to learning even more!
I dont need sudo. Just have weechat and wireguard preinstalled and im good.
Hi @Average4552!
I got up just now, 11:49 here (MST). Really late for me. Anyway, I received your Paypal payment, and in a second, I will go over to OneProvider and initiate reinstalling one of the servers to Debian 12. OneProvider's installs seem to take a while, sometimes two or three hours. Occasionally an install seems to fail and needs to be restarted. As soon as I get a working Debian install, I will add your key to the root account and let you know. Meanwhile, I'm here if you have any more questions or concerns. Thanks very much!
Best!
Tom
Edit to add: Install started at Mon 12 Feb 2024 12:09:25 PM MST
Hi again @uekix!
Thanks for your message! Please allow me some time to consider your request.
Best wishes and kindest regards,
Tom
Mon 12 Feb 2024 02:36:54 PM MST
Hi @Average4552!
Sent you login info via PM. Please let me know if you can get in. Please feel free to go ahead with awesome projects. Please keep in touch by posting here, PM, email, etc. If you want or need anything, just let me know.
Have fun!
Best!
Tom
Seems like the possibly old installer at OneProvider / Psychz might be making the /boot partition too small for newer Linuxes. We get a
No space left on device
error withapt autoremove
.apt autoremove
errorThe standard solution to this problem seems to be booting into a rescue system and increasing the size of the /boot partition.
What happens if we just remove /boot/initrd.img-6.1.0-13-amd64 and vmlinuz-6.1.0-13-amd64 or maybe change those two files to something smaller but with identical names? With the two files gone, or smaller,
apt
might have enough space to complete theautoremove
operation. Obviously not a long term solution, but might it be good enough for the remaining two weeks of the server contract period?Some other Debian servers at another provider are still on 6.1.0-13.
apt
on those other servers seems not trying to update to 6.1.0-18. Why is this server trying to update when the other servers are not?@Average4552
While i have no clue about the updates i can tell you that a boot partition of 108MB is going to be painful. Even 128MB isn't enough. I figure it should be at least 256MB (i'd rather play it safe and go straight for 512MB) unless you fancy running into regular errors during update-initramfs. It should be possible to sidestep the errors by moving files out during the update though if all you want is to keep it functioning for a short amount of time.
Setting
MODULES=dep
in/etc/initramfs-tools/initramfs.conf
usually helps a bit (at least on Ubuntu) to reduce to initrd image sizes.@Not_Oles I'm afk ATM but will have some more time to look into it this afternoon. If it were me, I'd
I mean it's running on .0-18 now, right, but wants to rebuild .0.13? Am I understanding that right?
Good morning! Happy Valentine's Day, LETizens! 🌹🌹🌹
If I have understood correctly, we're running 18 and it wants to rebuild 18. I guess 13 is still there so we can revert to it if there is a problem with 18.
First, I will try @cmeerw's suggestion. . . . Thanks @cmeerw!
Before changing anything, inside
/etc/initramfs-tools/initramfs.conf
, I am seeing:References:
https://bugs.launchpad.net/ubuntu/+source/initramfs-tools/+bug/2002230
https://bugs.launchpad.net/oem-priority/+bug/1842320
https://askubuntu.com/questions/299396/how-to-reduce-the-size-of-initrd-img-on-ubuntu-13-04
man uppdate-initramfs
Looks like, after setting
MODULES=dep
, the update commands to run might be:update-initramfs -u
and maybe also
update-initramfs -d -k 6.1.0-13-amd64
Let's try!
Wow! As shown below, the initrd image shrunk more than I imagined!
initrd.img-6.1.0-18-amd64
went from 30826674 bytes to 8086406 bytes. That might be enough extra space!Let's first see if the server will reboot successfully with the new, smaller initrd. And then, if it does reboot successfully, let's try the apt command again. . . .
Okay, back in after a successful reboot!
Trying
apt
againMmmm! Okay! Looks like we have a new initrd.img for 6.1.0-18 since the timestamp has changed:
-rw-r--r-- 1 root root 8086406 Feb 14 12:31 initrd.img-6.1.0-18-amd64 # <====
-rw-r--r-- 1 root root 8086422 Feb 14 12:51 initrd.img-6.1.0-18-amd64 # <===
Also, we have a little extra space in /boot
We still have the option of rebuilding the initrd.img of 6.1.0-13 to a smaller size or even eliminating it. But, for now, maybe leave it alone because it's less of an opportunity to revert if we change it.
Thanks again to @Average4552 for sharing the server and finding the issue. Thanks to @cmeerw for suggesting an excellent fix!
Ansible? Anyone use it? Anyone have alternatives? I'd like to try to set up an ansible playbook on this one.... BUT.... I really hate config files. They're scary, they take time to LEARN, and the whole thing takes less time to do stuff in the future, SOOO.... if not now, when?
I run 19 cloud servers, all but one running Tor relays, one of which is a Tor exit at Hostmaze (RO). I could have used Ansible much earlier to save some time in setting them all up. I like them all looking alike. All but one run Debian, the oddball runs Fedora Kinoite (just starting with that one, learning as I go), and I run EndeavourOS on my home machine. I wanted to try Fedora on the one b/c I wanted to try an immutable OS (Atomic as Fedora calls it), and from what I've read, Fedora as a rolling release breaks less than Arch. Plus I love the look of KDE on my home machine so that's why I went with Kinoite and not Silverblue. It also interests me how Kinoite and Silverblue are containerized - something I've not dabbled in very much. So we'll see with that.
@Not_Oles I set myself up a user account in the sudo group with the MetalVPS machine:
installed
ufw
:allowed OpenSSH through the firewall:
and set it up where I can log in as a user with my SSH key:
Then I left that terminal open while I tried to log in as user (in case I messed something up I wouldn't be locked out, I could fix it from the root terminal):
So since I was successful with that, I'm done with logging in as root on that. I feel safer and less likely to mess things up that way.
These are the first things I do with a new cloud server. I keep a LibreOffice document on the steps to take. Most help I got on this were through DigitalOcean's tutorials. They're step-by-step and explained well as to what's happening.
Happy Valentine's Day to all!
@Not_Oles Oh yeah I usually change my SSH port on my cloud machines to help quiet the noise from attempted bot attacks. Prohibiting password login is IMHO the best #1 thing you can do to harden your server against attacks. I didn't do it on this machine b/c I didn't want to lock you or anyone else out, but I run
$ shuf -i 49152-65535 -n 1
to give me the random port for SSH and then editsshd_config
thensudo systemctl restart ssh
. Weird how most distros need to restartssh
but Arch needssshd
. idkI also install
zsh
as my default shell with some cool plugins, install topgrade so I can check for upgrades on all my machines with one command from my local machine, and then installrsync
so I can backup my cloud servers to my home machine using luckyBackup.Maybe you might remove my keys from /home/michael/.ssh/authorized_keys ?
Probably tomorrow I also will make myself a Neighbor account.
@Average4552 Sorry, I might have forgotten to mention that password authentication already has been disabled.
Thanks for all the details of everything you have been doing to move in and to get the place comfy!
If you wanna change ssh to a random high port, please just go ahead and do so. Please PM me the new port.
After changing sshd_config or something in sshd_config.d, I always restarted sshd on every distro. I never restarted ssh. Now I have to study up a bit on the differences between restarting ssh and sshd.
Thanks! Have fun! Appreciate your sharing the cost of the server and all your detailed comments! Best!
Tom
Right now @Average4552 and I are the only folks on this machine. Anybody else want to come aboard? Anybody else for the FreeBSD machine? Also only one other guy besides me on the FreeBSD machine. Please note that these two machines are set to cancel on February 26. If desired, we might be able to get replacements direct from Psychz for a penny more per month, but with IPv6 in addition to IPv4.
On our Debian 12 machine here, I am seeing:
On Debian 12 at another provider, I am seeing the same, except s/remote-fs/paths :
On Fedora Rawhide there is no
ssh.service
:Hello!
I've been having some fun compiling the text editor, Elvis.
On FreeBSD 14.0-RELEASE, I see this error:
I also see
which I am guessing might be the needed library. If that's right, how come
ld
can't find it? I have to go study up onld
. Probablyman ld
will help. But, meanwhile, does anybody want to give me a hint about how to fix this error? Thanks!P.S. The error doesn't happen with Elvis 2.2.
maybe you need to install that package library?
I looked around a bit and didn't find what looked like an installation candidate.
Thanks for helping!
Oopsie! ipc vs icp.
Searching for libipc github brings up several results, the first of which is https://github.com/mheily/libipc.
However, searching the FreeBSD ports portal suggests that no libipc has been ported to FreeBSD: https://ports.freebsd.org/cgi/ports.cgi?query=libipc&stype=all&sektion=all says "Sorry, nothing found."
I heard that the missing library might be part of the X Window System. One option might be to install X and thus hopefully obtain the missing library. Alternatively, set a
configure
option to disable X in elvis.From the Makefile: