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.
Comments
Not for long, last time 2 hours, this time not sure even.
Maybe even less, Frankfurt be fucking tiny.
my 11 whatever o clock order still pending.
Be strong. There is always hope.
mai ohrder from 4 novembir 2025 (wow very far off date but i am patient) 8 27 39 still pendin 🤔. It s very long wait i am sad face 😢. need fast activation plzzzz. maybe neckst monf it will be done. i hope everyone have nice day tho sehr gut support team please look at this akount 🙏
...
Ah shit sorry, wrong thread.
My order from 4th nov 2025 at 8:27:39 still pending...
(thanks @brauni for the function I didn't know I wanted)
Since we’re waiting for new chickens to drop anyway, I was wondering — @OhJohn, do I remember correctly that you’re building a 3-node Proxmox cluster on those KS-LE-B servers? If so, may I ask how you handle high availability in terms of IP failover? As far as I know, floating IPs aren’t available for the ECO line. So, if one node goes down and a VM gets spun up on another, how do you deal with the IP change?
Sorry for the noob question, and thanks in advance for your help!
Has the dust settled?
I've been away for a while.
You were missed
Early birds waiting for some last orders to be delivered. Other people praying for an unlikely KS-LE-B restock next week.
On proxmox, if a node dies, a VM gets started again.
So a simple bash script for dynamic dns update is enough.
People still gambling on invoices, not yet.
Too busy knee deep in Cisco IOS and RouterOS understanding protocols (and why Cisco is a little special).
Actually not promox planned, but a seaweedfs cluster for s3 storage etc.
But in general, this might work with promox as well: I would put two+ small ovh vps in front as lb (with a failover ip and eg. keepalived) and haproxy to connect to the (e.g. promox) cluster internally (and with internal failover if you have redundancy in the cluster).
Not using promox though, so this is just general advice about typical ha setups, might be different or more special for promox.
Thanks for the explanation, but that's a bit out of my league
by now the function mostly feeds my depression.
Also why did you remove the dates of the paid orders?
Too brainy for me too
@raindog308, it seems that we already have the cover photography for the blog post, so interview of Octave on LEB wen?
cheaky fredo1664
hey @raindog308 do you still wear that belt?
Ah sorry it's my Linus phase... I removed it because it's the expiration date and I didn't think it was useful information for a pending delivery order, but it's an easy fix if you want it back. Or perhaps I can put the order date instead. Or both.
I believe they are.
Additional/Failover IPs are, but not Floating IPs. That makes it a bit harder, since you need to migrate the IP using the Interface or API yourself or by script on failover
Ah yes, years ago you had them only at SyS and (or only?) Rise, I think you can use them at KS servers now as well. But you can definitely use them on the new vps.
For automatic failover between Kimsufi server I use keepalived with check and notify scripts.
You can google yourself, but that was the first thing that popped up
https://www.tobru.ch/keepalived-check-and-notify-scripts/
So you can attach failover ips to KS servers now. Thought so.
And btw with failover ips GRA-RBX-SBG is the magic circle as you can move the failover ip between those DCs. That is the only magic one in OVH, all other DCs are intra-DC only.
See, and I love to follow your Epyc quest - most of that stuff is out of my league
plus vRack as it supports using the IPs arround the globe but with the gateway in one location
Yes, but having a vRack would be unintentional (leftover) luck with a KS. Normally KS is without vRack.
And having the gateway in just one location is not what I would like.
Instead of vRack I still use several isolated vpn meshes across DCs. But with that you can also go inter-provider.
That's true, was talking about additional ips in generel. As always, it depends on your specific use-case. There is always a way to do it, you just need to find it and commit to it
Is the HA a requirement or are you homelabbing/practicing? Because I don’t think HA is the use case they were thinking about when they started the Kimsufi brand. Althought I guess that nowadays Kimsufi are not what they use to be (shitboxes).