Case:
We want to disable IPv4 on an interface
Solution:
sudo ip -4 addr flush dev enp0s1
Explanation:
The above command removes all IPv4 addresses for interface enp0s1. Important, remember that this disabling is only temporary.
Sunday, October 29, 2023
Friday, October 13, 2023
How to uninstall brew in MAC
How to uninstall brew in MAC
Option 1:
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/uninstall.sh)"
Option 2: NONINTERACTIVE=1 /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/uninstall.sh)"
Tomado de: https://github.com/homebrew/install#uninstall-homebrew
NGINX Reverse Proxy for an IPv6-Only Server Farm
Introduction
This work presents a very simple way to offer dual-stack web access to an IPv6-only server farm using NGINX. The continued growth of the Internet and the gradual adoption of the IPv6 protocol means that it is essential to ensure connectivity and accessibility for clients using both IPv4 and IPv6. We will explain how to configure NGINX to support dual-stack web access, we will address how to configure NGINX as a reverse proxy that listens on both IPv4 and IPv6 addresses, as well as how to correctly route incoming requests to backend servers with only IPv6 addresses. By the way, among many other benefits, what we will discuss in the following article is an important step towards the preservation of IPv4 addresses.
What is a Reverse Proxy?
In [1], Cloudflare defines a Reverse Proxy Server as follows: “A reverse proxy is a server that sits in front of web servers and forwards client (e.g. web browser) requests to those web servers. Reverse proxies are typically implemented to help increase security, performance, and reliability. In order to better understand how a reverse proxy works and the benefits it can provide, let’s first define what a proxy server is.”
What is a Proxy Server
In [1], Cloudflare also provides the following definition for a proxy server: “A forward proxy, often called a proxy, proxy server, or web proxy, is a server that sits in front of a group of client machines. When those computers make requests to sites and services on the Internet, the proxy server intercepts those requests and then communicates with web servers on behalf of those clients, like a middleman.”
What are the benefits of a Reverse Proxy?
• A reverse proxy can offer IPv4 or transparent IPv6 to clients serviced from an IPv6- only server farm (which is what we will focus on). • Scalability: The use of a reverse proxy allows adding or removing backend servers as needed without affecting end users. This makes it easier for applications to scale out, allowing them to handle a larger number of concurrent users and requests. • Static content caching: NGINX can cache static content such as images, CSS files, and JavaScript, thus reducing the load on backend servers and increasing content delivery speed. This decreases page load times and the required bandwidth. • Security: NGINX acts as a point of entry to the application, providing an additional layer of security. It can perform functions such as request filtering, DDoS attack prevention, SQL injection protection, and client authentication. NGINX can also enable the use of SSL/TLS encryption for communication between clients and the backend server. • Advanced routing: A reverse proxy allows performing advanced routing based on various criteria, such as domain name, URL, or HTTP headers. This is useful when we need to direct traffic to different backend servers based on the specific attributes of the requests. • Consolidation of services: NGINX can act as a single point of entry for various backend services. This simplifies the infrastructure by consolidating multiple services on a single server, thus simplifying management and maintenance. • Enhanced performance: NGINX is lightweight and resource efficient by design. Its streamlined architecture and ability to handle large numbers of concurrent connections make it a popular choice for improving web app performance. • Load balancing: A reverse proxy such as NGINX can distribute incoming traffic across several backend servers. This helps balance the load and guarantees that no server is overloaded, which improves an application's performance and responsiveness.
Topology
What is our Goal Today?
The edge server (Reverse Proxy Server) will be able to receive IPv4 and IPv6 HTTP requests, and depending on the website a user wishes to visit (domain), will forward the request to the right server. This is what will happen in our example:
The client visits: The request is sent to:
server-a.com → 2001:db8:123::1
server-b.com → 2001:db8:123::2
server-c.com → 2001:db8:123::3
server-a.com → 2001:db8:123::101
server-b.com → 2001:db8:123::102
server-c.com → 2001:db8:123::103
Requirements
• Linux with NGINX on the Reverse Proxy Server • Super user access • Web server on each of the servers in the farm • IPv4 and IPv6 Internet connectivity • Internal IPv6 connectivity
Let's get started
Let's get started
1) Install NGINX in all servers #apt update #apt install nginx
2) Create the websites in the NGINX reverse proxy
File /etc/nginx/sites-available/server-a.com
server { listen 80; listen [::]:80;
server_name server-a.com;
location / {
proxy_pass http://[2001:db8:123::101]; }
}
File /etc/nginx/sites-available/server-b.com
server { listen 80; listen [::]:80; server_name server-b.com; location / { proxy_pass http://[2001:db8:123::102]; } }
Archivo /etc/nginx/sites-available/server-b.com
server {
listen 80;
listen [::]:80;
server_name server-b.com;
location / {
proxy_pass http://[2001:db8:123::102];
}
}
File /etc/nginx/sites-available/server-c.com
server {
listen 80;
listen [::]:80;
server_name server-c.com;
location / {
proxy_pass http://[2001:db8:123::103];
}
}
3) Create symbolic links to enable the sites configured above:
root@ProxyReverseSRV:/etc/nginx/sites-enabled# ln -s /etc/nginx/sitesavailable/server-a.com /etc/nginx/sites-enabled/server-a.com
root@ProxyReverseSRV:/etc/nginx/sites-enabled# ln -s /etc/nginx/sitesavailable/server-b.com /etc/nginx/sites-enabled/server-b.com
root@ProxyReverseSRV:/etc/nginx/sites-enabled# ln -s /etc/nginx/sitesavailable/server-c.com /etc/nginx/sites-enabled/server-c.com
4) Remember to restart NGINX
$sudo systemctl restart nginx
About the logs
Logs are extremely important for any company or ISP that wishes to review incoming connections.By default, NGINX will use its own IP address for outgoing connections, which results in the
loss of the address of the client that originated the HTTP request. But don't worry. NGINX
has the solution: proxy_set_header. This requires configuring both the end server and the
Reverse Proxy server.
1) On the Reverse Proxy Server, we must configure the website assets.
# Example of nginx reverse proxy that allows logging the
client's
# original address and port number
location /examples {
proxy_pass http://[2001:db8:123::103];
proxy_buffering off;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-Host $host;
proxy_set_header X-Forwarded-Port $server_port;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
}
2) On the end server, add the following in the http section of the /etc/nginx/nginx.conf
file:
set_real_ip_from 2001:db8:123::100; #replace the IP address
with that of the proxy
real_ip_header X-Forwarded-For;
real_ip_recursive on;
Example:
http {
…
set_real_ip_from 2001:db8:123::100;
real_ip_header X-Forwarded-For;
real_ip_recursive on;
…
}
With these settings, the receiving server will trust the X-Forwarded-For header set to
2001:db8:123::100 and will log the client's source IP to /var/log/nginx/access.log.
Summary
The proposed design allows managing a 100% IPv6-only web server farm with access to both the IPv4 and the IPv6 worlds in a very simple, scalable, and efficient manner. This results in various benefits, including having to manage only one TCP/IP stack, simplicity, security, and even saving IPv4 addresses.
References
• [1] https://www.cloudflare.com/es-es/learning/cdn/glossary/reverse-proxy/
• https://www.digitalocean.com/community/tutorials/how-to-configure-nginx-as-areverse-proxy-on-ubuntu-22-04
• GitHub. LACNIC Blog Post Help Files for the entire project:
https://github.com/LACNIC/BlogPostHelpFiles/tree/main/2023_Ofreciendo_conectivid
ad_Dual_Stack_a_servidores_Web_en_una_granja_de_servidores_100_IPv6_Only
Thursday, July 6, 2023
Google returns: 403. That's an error. Your client does not have permission to get URL / from this server. That's all we know.
Introduction:
You want to do a google search and the page returns: "403. That’s an error.
Your client does not have permission to get URL / from this server. That's all we know."
In my case I was using an IPv6 tunnel with Hurricane Electric, specifically the /64 that they deliver in the tunnels.
Solution?
Ask Hurricane Electric at the portal for a /48 routed. That's it! Then I removed the old /64 prefix from the router's SLAAC, leaving only a /64 belonging to the /48.
Good luck!
Thursday, June 29, 2023
Thursday, May 25, 2023
Strange ssh behavior on MAC - Copy/Paste Issues
Situation:
Strange behavior of SSH on MAC, problems with copy/paste in terminal during ssh. Does the clipboard work in other applications?
Solution:
At least in "vi" the solution is very simple. Edit the file: ~/.vimrc and paste the following content:
if !has("gui_running")
set mouse=
endif
Luck!
Thursday, March 16, 2023
A Look at LACNIC’s IPv6-Only Members
Introduction
This paper seeks to analyze the announcement and status of the prefixes held by organizations known as IPv6-only LACNIC members. These are organizations that have received IPv6 prefixes, that may or may not have received an autonomous system (ASN), and that have not been assigned an IPv4 prefix by LACNIC. The results of this analysis will help us improve our understanding of the uses and needs of our members across the region.
Information sources used in this analysis
The source of information used in this analysis was LACNIC whois, with data obtained throughout the month of January 2023. This information only includes those members who have been assigned IPv6 but not IPv4 resources by LACNIC. They may or may not have been assigned an autonomous system number.
Data processing
Python3 on a Jupyter notebook. We used the public APIs of LACNIC and RIPE NCC.
LACNIC WHOIS
RIPE NCC API
APNIC IPv6 Penetration by ASN
LACNIC’s delegated-extended File:
Findings
We found 483 organizations that are considered IPv6-only LACNIC members, which we divided into:
IPv6-only members with ASN: 343 members
IPv6-only members without ASN: 140 members
Results for the 483 IPv6-only members (with and without ASN):
An analysis of the 483 members allowed us to obtain the following information:
– 261 announce the entire or a portion of the prefix
– 208 announce the entire prefix they received from LACNIC
– 53 announce a portion of the prefix they received from LACNIC
– 222 do not announce the prefix
Results for the 343 IPv6-only members with at least one ASN:
A total of 343 of the 483 IPv6-only members have at least one ASN (71%).
Out of the 343 members that have an ASN:
163 announce the prefix (47.52%)
120 announce the entire IPv6 prefix (73.61%)
43 announce a portion of the IPv6 prefix (26.38%)
180 do not announce the IPv6 prefix (52.48%)
Results for the 140 IPv6-only members without an ASN:
A total of 140 of the 483 IPv6-only members do not have an ASN (29%).
Out of these 140 IPv6-only members with no ASN:
98 announce the prefix (70%)
82 announce the entire prefix (83.67%)
16 announce a portion of the prefix (16.32%)
42 do not announce the prefix (30%)
What can we conclude from the charts above?
The first thing that stands out is that IPv6-only members without an ASN have 23% more announcements than LACNIC members who do have an ASN. In other words, these members have asked another organization to announce their prefix.
In both cases, the number of unannounced prefixes is very high, and it might be interesting to find out whether there is a reason for this.
It is particularly striking that the percentage of partial announcements is almost identical (11% of members with an ASN and 12% of members without an ASN).
Trying to identify whether the ASNs have IPv6 traffic
Knowing that there are 343 ASNs with assigned IPv6 prefixes and considering that these are also IPv6-only members, one can presume that they must have “medium high” IPv6 traffic. Therefore, we analyzed each autonomous system to determine its IPv6 traffic on 23 January 2023.
How do we find out if an ASN has IPv6 traffic?
As many of you know, APNIC has been measuring IPv6 traffic for each ASN for many years. You can learn more about these studies here: https://stats.labs.apnic.net/ipv6
Based on the above, we wanted to find out if the known ASNs of IPv6-only LACNIC members actually had IPv6 traffic (beyond announcing their prefix).
Unfortunately, we were unable to obtain information for 274 (79.88%) of the 343 autonomous systems that were analyzed. However, it is important to note that this does not necessarily mean that they have not deployed IPv6, but that the traffic generated is very low and does not appear in the measurements by APNIC. A total of 32 ASNs were reported with 0% IPv6 traffic and 1 ASN with 88% IPv6 traffic.
Are our IPv6-only members actually IPv6-only?
Finally, we wanted to find out if our members (IPv6-only with an ASN) are truly as IPv6-only as their name implies.
For this particular case, we relied on the RIPE NCC API to obtain the information presented below.
Out of the 343 IPv6-only members with at least one ASN:
– A total of 540 prefixes were being announced (including v4 and v6)
IPv4 announcements: 220
IPv6 announcements: 320
In IPv4 announcements, the average prefix length is: 23.56
In IPv6 announcements, the average prefix length is: 38.09
– 66 members “became” DualStack, i.e., the ASN announces both IPv4 and IPv6
– 90 members announce IPv6 only
– 27 members announce IPv4 only
– 160 do not announce any prefix
And to which RIR do the IPv4 prefixes announced by IPv6-only LACNIC members belong?
LACNIC ARIN RIPE NCC AFRINIC APNIC
28 166 17 9 0
Sankey Diagram – A Look at LACNIC’s IPv6-Only Members
Conclusions
Results suggest that, while a significant number of LACNIC members consider themselves to be IPv6-only, we observed that more than 50% of these members have not started announcing their IPv6 prefix. At the same time, we noticed that many of those who have deployed IPv6 continue to use IPv4. This means that, although in recent years IPv6 adoption has grown in the region, there is still a long way to go to achieve widespread IPv6 deployment. In any case, it is important to continue to monitor the adoption of the new protocol.
Finally, being an IPv6-only LACNIC member also allows an organization to participate in the Internet ecosystem.
Subscribe to:
Posts (Atom)