Merge branch 'master' of code.ungleich.ch:ungleich-public/ungleich-staticcms
This commit is contained in:
sanghee
commit
f4a70a64e8
3 changed files with 134 additions and 7 deletions
127
content/u/blog/2022-04-01-infrastructure-update/contents.lr
Normal file
127
content/u/blog/2022-04-01-infrastructure-update/contents.lr
Normal file
|
|
@ -0,0 +1,127 @@
|
|||
title: URGENT: Infrastructure update Announcement
|
||||
---
|
||||
pub_date: 2022-04-01
|
||||
---
|
||||
author: ungleich
|
||||
---
|
||||
twitter_handle: ungleich
|
||||
---
|
||||
_hidden: no
|
||||
---
|
||||
_discoverable: yes
|
||||
---
|
||||
abstract:
|
||||
We are updating our infrastructure - what actions do you need to take
|
||||
---
|
||||
body:
|
||||
|
||||
## APRIL's FOOLS DAY
|
||||
|
||||
None of the news below is true - this was just part of our 2022
|
||||
April's fools day contribution.
|
||||
|
||||
## Urgent release: ungleich infrastructure update
|
||||
|
||||
To the attention to every ungleich customer: today we are updating our
|
||||
infrastructure to the latest technologies. All changes will be applied
|
||||
at 23:42 CEST today.
|
||||
|
||||
What does that mean to you? The following sections answer this in detail.
|
||||
|
||||
## Network upgrades: Turning IPv4 off
|
||||
|
||||
As of today all IPv4 traffic will be disabled. IPv4 has become
|
||||
too expensive and it is not worth keeping it online.
|
||||
|
||||
The migration will be as follows: every IPv4 address like A.B.C.D will be
|
||||
translated to an IPv6 address like 2a0a:e5c0:b00::A.B.C.D.
|
||||
|
||||
As a bonus, every customer will get their own "cafe" IPv6 network.
|
||||
|
||||
Additionally every server will be connected also via
|
||||
[SCION](https://scion-architecture.net/), the new Internet.
|
||||
|
||||
## Network upgrade: Removal of old connections: no more 2G/3G/4G
|
||||
|
||||
You may be aware that slow Internet connections are draining server
|
||||
resources: long lasting connections without much data transfer is a
|
||||
burden for every server. For that reason, we will disable access to
|
||||
our data center from any 2G/3G/4G mobile connection.
|
||||
|
||||
You are required to use 5G, 6G, 7G, Fiber or 56k modems to access Data Center
|
||||
Light. The latter was included as a fallback, in case you don't have
|
||||
support for a modern Internet connection.
|
||||
|
||||
|
||||
## Storage migration
|
||||
|
||||
Our storage is so far based on Ceph, a distributed storage. To follow
|
||||
modern standards, we are migrating all our data to the blockchain.
|
||||
|
||||
To avoid vendor logins and possible shortcomes of a particular
|
||||
blockchain, we are introducting the "ungleich blockchain", which
|
||||
stores data on **all** available blockhains.
|
||||
|
||||
If you are worried about privacy implications of all data being
|
||||
public, don't worry: we are encrypting all data using an
|
||||
[Enigma machine](https://en.wikipedia.org/wiki/Enigma_machine).
|
||||
|
||||
Optionally, on request, we can also encrypt your data as NFTs.
|
||||
|
||||
In regards to performance, we expect a huge performance increase, as
|
||||
all data will be saved everywhere.
|
||||
|
||||
## Computing power
|
||||
|
||||
As you might have read recently, we are investigating heating houses
|
||||
with the server heat. This project is now in productive state and we
|
||||
require much more computing power.
|
||||
|
||||
For that reason, at the same time today, our whole server
|
||||
infrastructure will be replaced with [quantum
|
||||
computers](https://en.wikipedia.org/wiki/Timeline_of_quantum_computing_and_communication).
|
||||
|
||||
As qubits, the equivalent of bits on quantum computers, operate on
|
||||
probabilities, all calculations for our customers are not guaranteed
|
||||
anymore, but instead they will be executed by a certain probability.
|
||||
|
||||
The exact probability is derived from your society behaviour
|
||||
points. If you are [in the good
|
||||
place](https://en.wikipedia.org/wiki/The_Good_Place) you get an
|
||||
additional uncertainty factor of 3.14.
|
||||
|
||||
## Removal of Kubernetes
|
||||
|
||||
We recently introduced [kubernetes](https://kubernetes.io/) in our
|
||||
infrastructure. After only 2 years of R&D we will have to shutdown all
|
||||
Kubernetes based clusters. The main reason is that the actual
|
||||
complexity of Kubernetes systems is **not high enough**. Plainly
|
||||
speaking, our engineers are bored. For that reason:
|
||||
|
||||
We will replace Kubernetes with a [home
|
||||
made](https://en.wikipedia.org/wiki/Don%27t_repeat_yourself) solution
|
||||
that consists solely of a manually managed cluster of Windows
|
||||
Servers. Microsoft is a direct partner in this and provides us with
|
||||
a new Windows version: **Microsoft Windows Quantum ME 3.11**.
|
||||
We expect great stability from this release with a minimum complexity
|
||||
factor of 1000x added.
|
||||
|
||||
## Maintenance and Customer support
|
||||
|
||||
We have always been very community orientated. Today we wanted to take
|
||||
the next big step and we turn over the customer support and system
|
||||
maintenance to the community. What does that mean specifically?
|
||||
|
||||
Anyone interested in maintaining our infrastructure can get full root
|
||||
access on all systems.
|
||||
|
||||
As everyone now has full root access, the customer support is also
|
||||
shifted to a decentralised, [bittorrent
|
||||
based](https://en.wikipedia.org/wiki/BitTorrent) support community.
|
||||
|
||||
If you have any questions to the recent changes, please connect the
|
||||
decentrally organised support community. Unfortunately, due do the
|
||||
decentralisation, we cannot offer a single contact reference.
|
||||
|
||||
However the Web3 advocate of your choice will certainly be able to
|
||||
provide you answers.
|
||||
|
|
@ -193,8 +193,9 @@ CoreDNS inside kubernetes is by default configured to allow resolving
|
|||
for *any* client that can reach it. Thus if you make your kube-dns
|
||||
service world reachable, you also turn it into an open resolver.
|
||||
|
||||
At the time of writing this blog article, the following coredns
|
||||
configuration **does NOT** correctly block requests:
|
||||
The following coredns configuration **does** correctly block
|
||||
requests, **IF your coredns version is new enough**:
|
||||
|
||||
|
||||
```
|
||||
Corefile: |
|
||||
|
|
@ -212,10 +213,9 @@ configuration **does NOT** correctly block requests:
|
|||
...
|
||||
```
|
||||
|
||||
Until this is solved, we recommend to place a firewall before your
|
||||
public kube-dns service to only allow requests from the forwarding DNS
|
||||
servers.
|
||||
|
||||
We tested this with
|
||||
[coredns-1.8.4](https://github.com/coredns/coredns/issues/4697) in
|
||||
which the ACL behaviour is fixed.
|
||||
|
||||
## More of this
|
||||
|
||||
|
|
|
|||
|
|
@ -14,7 +14,7 @@ aspect was the main reason to cancel the event, the logistic aspects
|
|||
days after departure) also played a big role for us to cancel the
|
||||
event.
|
||||
|
||||
## Introducting the Mini-Hack4Glarus
|
||||
## Introducing the Mini-Hack4Glarus
|
||||
|
||||
The measures are slowly being removed in many countries, including
|
||||
Switzerland. While we are taking it a bit slower at ungleich in
|
||||
|
|
|
|||
Loading…
Reference in a new issue