Ansible through Ubuntu (WSL) on Windows 10

Windows Subsystem for Linux (WSL) allows you to run Linux straight from your Windows Desktop. I use this on a daily basis for running Ansible scripts without having to install VM’s. Make sure you installed al latest updates.

Enable WSL feature

Open up a Powershell box as Administrator (search powershell, right click and run as Administrator).

Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Windows-Subsystem-Linux

This will initiate the installation and once finished ask if you would like to reboot your system. Go ahead and do that. When the reboot is done search for ‘bash’ and open that, it will first require a few anwsers. Simply fill out all the questions and once that is done you will have Ubuntu up and running.

Install Ansible

Now you are basicly in a Linux environment so you can install Ansible the typical way. Again, in the ‘bash’ window of course, use these instructions:

sudo apt-get -y install python-pip python-dev libffi-dev libssl-dev
sudo pip install ansible

Should you get any permission errors (i did not have this time, but given the nature of how WSL works that could happen) install pip with the –user flag. This will cause it to install ansible in the users home dir, not globally.

You are done. Using the following command you can check what ansible version is now installed:

ansible --version

If you need the most recent version check out my other post here.

Install latest version ansible on Ubuntu 16.04 / 18.04

Ubuntu doesn’t ship with the newest version of ansible out of the box, sadly. You have to manually configure the PPA on your system in order to upgrade to the stable version. Follow these commands to install the PPA:

$ sudo apt update
$ sudo apt upgrade
$ sudo apt install software-properties-common
$ sudo apt-add-repository ppa:ansible/ansible

Hit Enter when asked, and once the process is done update your apt repos:

$ sudo apt update

Now you can either upgrade or simply install ansible:

$ sudo apt install ansible

This should be all, use the following to verify the ansible version:

$ ansible --version

Ubuntu 18.04 resize/expand (root) filesystem

Running Ubuntu 18.04 i ran out of disk space on my main partition. I increased disk space in VMware and needed to expand the partitions from within Ubuntu. Start with scanning for changes on your disk

echo 1 > /sys/class/block/sda/device/rescan

Verify that you can see the new (correct) disk space using:

fdisk -l

Create a new partition using “cfdisk”, navigate to the free space and hit “new”. After that hit Write to make sure the partition table gets written. Close cfdisk and either reboot or rescan to update your partition table. Now it’s time to add the disk space. First, find the new partition number:

fdisk -l

In my case sda3 was created to i’m going to create a new volume on it

pvcreate /dev/sda3

now extent my volume group with the newly added volume:

vgextend ubuntu-dev-box /dev/sda3

Extend the volume with all available (new) disk space:

lvextend -l+100%FREE /dev/ubuntu-dev-box/root

Now resize the filesystem:

resize2fs /dev/mapper/ubuntu--dev--box--root

SSH Tunnel to watch Netflix

I often use a ‘hopping server’ when connecting to clients, that means i need to login twice each time. To make my life easier i sometimes use an SSH tunnel so i can connect to clients directly.

SSH Tunnel can also be usefull when your office blocks netflix 😉

Local Port Forwarding

This will allow you to access remote servers direcly from your local computer. Let’s assume you want to use RDP (3389) to a clients hosts (10.0.1.1) and your hopping server is ‘hopping.server’

ssh -L 6000:10.0.1.1:3389 wieger@hopping.server

Now you can open Remote Desktop and connect to ‘localhost:6000’, directing you through the tunnel!

Remote Port Forwarding

This will make your local service/port acccessible from a remote host. Sometimes i use this to keep a ‘backdoor’ and login remotely (home or whatever).

Let’s say you want to make a webapplication (TCP 443) availible at port 6000 on the remote SSH server

ssh -R 6000:localhost:443 wieger@bontekoe.technology

Now you should be able to connect to port 6000 on the remote host (bontekoe.technology)

Dynamic Forwarding (Proxy)

This is ideal for people who want to use the internet safely/anonymous or for offices where Netflix is blocked 😉

Use a remote server to tunnel all web traffic (eg. home server), connect through SSH to it using the -D flag

ssh -D 6000 wieger@bontekoe.technology

Now open up your browser settings, navigate to the connection properties and enter a Proxy server (manually using SOCKS). Use 127.0.0.1 as host and 6000 as port. The tunnel will remain open as long as you are connected through SSH.

mod_pagespeed Module on Ubuntu 18.04

mod_pagespeed is an open-source Apache module created by Google to help Make the Web Faster by rewriting web pages to reduce latency and bandwidth. mod_pagespeed releases are available as precompiled linux packages or as source. (See Release Notes for information about bugs fixed)

Installation

  1. Update system

    apt update -y
    apt upgrade -y

  2. Install Apache

    apt-get install apache2 -y

  3. Enable Apache Startup

    systemctl start apache2
    systemctl enable apache2

  4. Install mod_pagespeed

    wget https://dl-ssl.google.com/dl/linux/direct/mod-pagespeed-stable_current_amd64.deb
    dpkg -i mod-pagespeed-stable_current_amd64.deb
    systemctl restart apache2

  5. Verify mod_pagespeed is running

    curl -D- localhost | head | grep pagespeed

Web Interface

mod_pagespeed has a very simple web-interface to see statistics. If you do not case, skip this step.

nano /etc/apache2/mods-available/pagespeed.conf

Add these lines to it:

<Location /ps_admin>
    Order allow,deny
    Allow from localhost
    Allow from 127.0.0.1
    Allow from all
    SetHandler ps-admin
</Location>

<Location /ps_global_admin>
    Order allow,deny
    Allow from localhost
    Allow from 127.0.0.1
    Allow from all
    SetHandler ps_global_admin
</Location>

After restarting apache you can go to http://<your url>/ps_admin

BGP Route-Leak causes a significant shift in European internet traffic to Chinese Backbone network

A series of unfortunate configuration mistakes?

Last Thursday, June 6, 2019 at 12:00 a.m. Dutch time, a two-hour outage began at various customers. Both business customers for fiber optic internet as well as private internet connections. At first, the extent of the disruption was not yet completely clear.

After several reports from customers I noticed that also on Dutch websites (such as allestoringen.nl) the complaints came in from people with connection problems. It soon turned out to be a national outage that primarily affected KPN (and indirectly many other parties including national payment systems).

A first traceroute showed that the traffic between Haarlem and Amsterdam went through different (completely illogical) paths. This had all the symptoms of a BGP hijack, The Dutch website Tweakers.net also mentioned this and expressed this suspicion.

Doing a bit more digging on the route my traffic was taking, i saw that my VPN connecting between Haarlem and Amsterdam was suddenly taking a de-tour through AS4134, known as the Chinanet Backbone Network.

For two hours, much of Europe’s Internet traffic passed through Chinese networks

This incident occurred after a Swiss hosting provider (Safe Host SA) started leaking more then 70.000 faulty routes to the Chinese backbone network. This Chinese network, in turn, forwarded these IP address announcements as valid to various large Tier-1 internet providers causing a huge traffic shift toward the Chinese Backbone network. The incident caused huge impact on the networks of Swisscom (AS3303), KPN (AS1130), Bouygues Telecom (AS5410) and SFR (AS21502).

Error or intention?

The official statement is; a configuration error by the Swiss hoster SafeHost. The remarkable thing is that the incorrectly advertised ranges were smaller and more specific then the ones advertised legitimatly. Some websites mention this could indicate the use of route optimizers but Safe Host SA confirms on twitter that they do not use bgp optimalisation software.

Now that Safe Host itself indicates that it is not the use of route optimizers that caused this disruption, the question arises how it is possible that such specific routes have been (wrongly) advertised. And why only to Chinanet Backbone? Safe Host SA is still investigating and not responding to questions on Twitter.

I am very curious about the explanation of this incident, if it ever comes.

Worldwide concern

Chris C. Demchak and Yuval Shavit released a publication in 2018 entitled “The Hidden Story of China Telecom’s BGP Hijacking”. In this publication they write how China has been able to divert specific traffic through Chinese POPs several times via BGP hijacks and the possible implications this has.

They also describe their concerns about the large (infrastructural) network presence of the Chinese Backbone Network in America (which is no different in the EU) while no other international network has similar presence in China. This large presence is one of the things that makes it easy for the Chinese to conduct a BGP Hijack on large scale and at the same time protecting their own infrastructure.

Using these numerous PoPs, CT (China Telecom) has already relatively seamlessly hijacked domestic US and cross US traffic and redirected it to China over days, weeks, and months as demonstrated in the examples below. The patterns of traffic revealed in traceroute research suggest repetitive IP hijack attacks committed by China Telecom.

The report argues that the Chinese government is using local ISPs for intelligence gathering by systematically hijacking BGP routes to reroute western traffic through its country, where it can log it for later analysis and provides the following examples;

  • Starting from February 2016 and for about 6 months, routes from Canada to Korean government sites were hijacked by China Telecom and routed through China
  • On October 2016, traffic from several locations in the USA to a large Anglo-American bank headquarters in Milan, Italy was hijacked by China Telecom to China
  • Traffic from Sweden and Norway to the Japanese network of a large American news organization was hijacked to China for about 6 weeks in April/May 2017.
  • Traffic to the mail server (and other IP addresses) of a large financial company in Thailand was hijacked several times during April, May, and July 2017.

Doug Madory from Oracle confirmed that AS4134 was redirecting traffic in a blogpost posted on 5th of november 2018.

In this blog post, I don’t intend to address the paper’s claims around the motivations of these actions. However, there is truth to the assertion that China Telecom (whether intentionally or not) has misdirected internet traffic (including out of the United States) in recent years. I know because I expended a great deal of effort to stop it in 2017.

How to prevent/secure this?

The problem with incidents like this is that the internet is running on the BGP protocol. BGP is a global protocol running between organizations and country’s crossing international borders. There is no single centralized authority, just internet providers that collaborate based on trust. The internet is intended to be open, transparant and so all ISP’s are trusted to play nice. Furthermore, there are initiatives that can improve the security of BGP in general, but these must be introduced on a large scale.

On the 12th of June, the Dutch National Cyber Security Center issued a report in which it once again shows how great the (digital) risks are for the Dutch society. In this report they also specifically mention the influence of China (among others).

“Countries such as China, Iran and Russia have offensive cyber programs against the Netherlands. This means that these countries are digital use resources to achieve both geopolitical and economic objectives to be achieved at the expense of Dutch interests “

It is time to immediately acknowledge and address these risks.

Test internet speed using speedtest-cli

Speedtest-cli is a great tool to test your internet speed using the Speedtest servers, make sure you have Python installed before installing Speedtest-cli.

Installing and Using Speedtest-CLI

  1. Update APT and install packages

    apt-get update; apt-get install python-pip speedtest-cli

  2. Test your speed!

    speedtest-cli
    Testing download speed........................................
    Download: 913.12 Mbit/s
    Testing upload speed..................................................
    Upload: 524.12 Mbit/s

  3. Share your speed 🙂

    speedtest-cli --share
    This will provide you with an image to share proving the speed.

Load Balancing Remote Desktop

Using HAProxy to loadbalance between RDS servers is usefull if you have more then one RDS servers and want users to connect to a single IP.

  1. Install Haproxy

    sudo apt-get update
    sudo apt-get install haproxy

  2. Add the RDP VIP (virtual IP) and RDP hosts

    defaults
    clitimeout 1h
    srvtimeout 1h
    listen VIP1 193.x.x.x:3389
    mode tcp
    tcp-request inspect-delay 5s
    tcp-request content accept if RDP_COOKIE
    persist rdp-cookie
    balance rdp-cookie
    option tcpka
    option tcplog
    server win2k19-A 192.168.10.5:3389 weight 10 check inter 2000 rise 2 fall 3
    server win2k19-B 192.168.10.6:3389 weight 10 check inter 2000 rise 2 fall 3
    option redispatch

Now we have HAProxy running on a 193.x.x.x ip address, when you connect to that IP it will direct you to one of the Windows 2019 machines. If one dies, it will remove it and you can reconnect to the last one that is online.

Installing phpIPAM

phpipam is an open-source web IP address management application (IPAM). Its goal is to provide light, modern and useful IP address management. It has lots of features and can be integrated with PowerDNS!

  1. Install Apache, PHP 7.2, MariaDB and GIT client

    apt-get install apache2 mariadb-server php7.2 libapache2-mod-php7.2 php7.2-curl php7.2-mysql php7.2-curl php7.2-gd php7.2-intl php-pear php7.2-imap php-memcache php7.2-pspell php7.2-recode php7.2-tidy php7.2-xmlrpc php7.2-mbstring php-gettext php7.2-gmp php7.2-json php7.2-xml git wget -y

  2. Run the secure installation for MariaDB

    mysql_secure_installation
    Enter current password for root (enter for none):
    Set root password? [Y/n]: N
    Remove anonymous users? [Y/n]: Y
    Disallow root login remotely? [Y/n]: Y
    Remove test database and access to it? [Y/n]: N
    Reload privilege tables now? [Y/n]: Y

  3. Create database and user

    MariaDB [(none)]> create database phpipam;
    MariaDB [(none)]> grant all on phpipam.* to phpipam@localhost identified by 'bontekoe123';
    MariaDB [(none)]> FLUSH PRIVILEGES;
    MariaDB [(none)]> EXIT;

  4. Clone it from Github

    cd /var/www/html
    git clone https://github.com/phpipam/phpipam.git /var/www/html/phpipam/
    cd /var/www/html/phpipam
    git checkout 1.3
    git submodule update --init --recursive

  5. Edit the config file

    cp config.dist.php config.php
    nano config.php
    Add your MariaDB database, username and password there.

  6. Import PHP IPADM Database

    mysql -u root -p phpipam < db/SCHEMA.sql

  7. Set the correct permissions

    chown -R www-data:www-data /var/www/html/phpipam
    chmod -R 755 /var/www/html/phpipam

  8. Apache configuration

    Either you use the default apache configuration or you can create a virtual host specificly for this. That i leave up to you.

  9. Finish the installer

    Go to http://x.x.x.x/phpipam and you will see the wizard, asking you to finish the installation. Follow the easy steps. After that you can login with admin/admin and change the default password.

  10. Create your first subnet

    Here you can create your first subnet. If can automaticly help you by discovery, checking hosts status and more.

Ceph Infrastructure Planning

Ceph is open source storage meant to facilitate highly scalable object, block and file-based storage. Because of it’s open source nature it can run on any type of hardware. The CRUSH algorithm will balance and replicate all data throughout the cluster.

At first i was running Ceph on regular servers (HP Proliant, DELL) but since +/- 2 years i decided to play around with custom configurations hoping to squeeze more iops out of it. As i found out, big part of getting the most out of Ceph is: network. My last Ceph configuration was so fast it took down all our servers 🙁

What was i doing?

In order to get the most out of ceph i created a setup using only NVMe drives. Here is my configuration for my ceph nodes:

Supermicro SYS-1028U-TN10RT (link)
Intel E5-2640v4 (10 Core, 2.4Ghz)
32GB DDR4-SDRAM
1x Intel SSD DC P4600
1x ASUS M.2 Hyper x16 Card
4x Samsung Pro 960 NVMe
1x Intel X520-DA2 Dual 10Gbe

I am using the Intel SSD DC P4600 as main (fast) storage and Samsung Pro NVMe’s for slightly slower storage. Everything was looking dandy.

So what happend?

One of our clients woke up and thought, “hey, let’s migrate some stuff today”. Generating only 1Gbps network traffic on the uplinks to the internet everything was looking fine. Until 15 minutes later everything went offline. All VM’s stopped and storage was down. Turns out that the client was migrating a set of DB’s to our network, after the scp copy stopped they extracted the files and imported them in the database server generating a lot of IO on that virtual server. That virtual server was however part of a database cluster so replicated all that data to 3 other VM’s.

As Ceph wrote the data to disk, it also replicated it across the Ceph pool and through the network generating way to much (20 Gbps per Ceph host) traffic on the access layer. Ceph was so fast, it took down the entire network.

How to resolve

As storage (specially Ceph 😉 ) is getting faster, network topology needs to change in order to handle the traffic. A design that is becoming very popular is the leaf-spine design.

This topology replaces the well known access/aggregation/core design and using more bandwidth on the connections to the spines we can handle much more traffic.

A 10 Gbit network is at minimum recommended for Ceph. 1 Gbit will work however the latency and bandwidth limitations will be almost unacceptable for a production network. 10 Gbit is a significant improvement for the latency. Also recommended is that each Ceph node is connected to two different leafs for external traffic (clients/vm’s) and internal traffic (replication).