Welcome to NBSoftSolutions, home of the software development company and
writings of its main developer: Nick Babcock. If you would like to contact
NBSoftSolutions, please see the Contact section of the
With the rise of privacy consciousness, people are looking to solutions like a
hosted VPN (I hear one should never use a free service), or self hosted like
algo. How does a VPN (a remote access
VPN – not a site-to-site VPN for the pedantic) help maintain privacy?
In the scenario of maintaining privacy or getting around geographic content
blocking, the VPN connects you to a server, oftentimes in a different country,
where it forwards all your traffic to the intended recipient. The recipient
responds to the server, which dutifully forwards back to you. So, if you live
in the US, but are VPNed into a German server and request content from India,
India will think you’re in Germany (this assuming countries have thoughts </joke>).
I’m going to show how to self host WireGuard,
which bills itself as easier to configure than IPSec and OpenVPN, while being
faster and more powerful. WireGuard is a component feature of of
streisand, but since we’re going to be
dealing with only a linux client and server setup we cut out the streisand
middleman and just use WireGuard. Theoretically, this cuts down on the bloat
and attack surfaces that are inherent with the wide array of software that
streisand installs (streisand is planning on supporting modular install in the
It should be noted:
WireGuard is not yet complete. You should not rely on this code. It has not
undergone proper degrees of security auditing and the protocol is still
subject to change.
This demonstration will be on a DigitalOcean Ubuntu 16.04 box, but it should be
easily adaptable for other platforms (as long as they are linux based).
The following script is to be executed on one’s server. This script will be
subsequently broken down.
Simple enough, how does this work?
Sysctl allow modifying kernel parameters at runtime, so here we’re allowing the
kernel to forward packets from one network interface to another. We need this,
as wireguard works by creating the VPN on another network interface (commonly
called wg0 or wgnet0). This interface, by itself, does not have internet
access, but with ip forwarding we can foward traffic from the VPN to the
interface that can communicate with the internet.
Forwarding is only important for the server because once connected to the VPN
the default client interface won’t be used anymore.
sysctl -w is only for changing kernel parameters at runtime. To persist these
settings, edit the relevant line in /etc/sysctl.conf
These commands fetches the latest wireguard version and installs it. Since
WireGuard hooks into the kernel, it attempts to automatically detect the
correct kernel to hook into. This should work flawlessly.
The one problem I’ve had is that for DigitalOcean controls the kernel through
their web interface (one can use a custom kernel but that is outside of the
scope of this post). Anyways, if you had tried to install a custom kernel ontop
of the one in DigitalOcean, wireguard will skip the correct kernel as it
believes it’s chrooted. Sorry for the tangent, but since I experienced this
problem, I figured I should document it for others.
Both the client and server need to generate a pair of keys. The server does not
need to know the client’s private key and vice versa; however they do need to
know each other’s public key to permit only authorized use of the VPN (else
anyone who knew your VPN server’s address could use your VPN).
When clients connect to the server, they can communicate directly with by using
the 10.192.122.1. We know that 10.192.122.1 can’t possibly be an internet
facing box because it falls under a private
network. The /24 is a CIDR
subnet mask that states that
this VPN will is capable of housing 254 clients. WireGuard then listens on port
51820 for interested clients.
Probably the most convulted section of the config, yet this step must not be skipped. This is how those lines configure the firewall.
When the VPN is created:
We accept packets from our VPN interface for packets being routed through the box
Then whenever a new connection is created (eg. our client wants to access
google.com, so the server needs to connect to google.com now), the outgoing
packets are altered to have the server IP address so google.com responds to
the server, which relays it back to the client.
Then when the VPN is destroyed everything in our firewall is deleted.
If you forget those lines, when you go to connect as a client your requests
will blackhole and it may appear as if you lost internet connection.
The peer section is for client information. The client that connects with the
given client public key is assigned 10.192.122.2 for their IP address.
The server is all setup so what does the client configuration look like?
Breaking down the config:
The 10.192.122.2/32 matches the same address as the server.
We set the client’s DNS server to that of the VPN server. This is not needed,
but I recommend it, as you want to communicate with servers (eg. google.com)
that are closest to your VPN server to minimize latency. For instance, if you
live in the US, VPNed into Singapore, and wanted google.com, you’d want to talk
to Singapore Google server (and not the one’s in the US) so that packets travel
the least distance.
It is actually possible to create and destroy VPN boxes on demand for next level privacy. Here is how one would do this:
Create DigitalOcean box by hand using the previous instructions.
Verify that the VPN works (wg-quick up wgnet0 on both client and server).
Using the DigitalOcean cli one can create a new server from our snapshot (so the server will have the same public and private key, it’ll just have a new ip address)
Update your client config to reference server’s new IP address
And just like that you can create and destory VPNs all around the world in under a minute.
I’m building a home server and I’m using Pi-hole to blackhole ad server domains in a test environment. It’s not perfect, but works well enough. The Pi-hole admin interface shows a dashboard like the following to everyone (admin users get an even greater breakdown, but we’ll focus on just the depicted dashboard)
It’s a nice looking dashboard; however, it’s a standalone dashboard that doesn’t fit in with my grafana dashboards. I don’t want to go to multiple webpages to know the health of the system especially since I want to know two things: is pi-hole servicing DNS requests and are ads being blocked. This post will walk through exporting this data from pi-hole into Graphite, which is my time series database of choice.
There is a blog post that contains a python script that will do this process. Not to pick on the author (they inspired me) but there are a couple of things that can be improved:
The python script is not standalone. It requires the requests library. While ubiquitous, the library is not in the standard library and will require either a virtual environment or installation into the default python
Any error will crash the program (eg connection refused, different json response, etc) and it’ll have to be manually restarted
No logs to know if the program is running or why it would have crashed
Does not start on computer reboot
Let’s start with making a network request:
Ah, let’s turn to jq to massage this data, which will, by default, prettyify the output
We somehow need to get the previous data into the <path> <value> <timestamp>format for carbon with lines seperated by newlines.
Since jq prefers working with arrays, we’ll transform the object into an array: jq 'to_entries'
Now we’re going to transform each element of the array into a string of $key $value with jq 'to_entries | map(.key + " " + (.value | tostring))'. Value is numeric and had to converted into a string.
Finally, unwrap the array and string with a jq -r '... | .' to get:
We’re close to our desired format. All that is left is an awk oneliner:
So what does our command look like?
Is this still considered a one-liner at this point?
I’ve add some commandline options to curl so that all non-200 status codes are
errors and that curl will retry 5 times up to about a half a minute to let the
applications finish booting.
We could just stick this in cron and call it a day, but we can do better.
systemd allows for some nice controls over our script that will solve the rest
of the pain points with the python script. One of those pain points is logging.
It would be nice to log the response sent back from the API so we’ll know what
fields were added or modified. Since our script doesn’t output anything, we’ll
capture the curl output and log that (see final script to see modification, but
With that prepped, let’s create /etc/systemd/system/pihole-export.service
Type=oneshot: great for scripts that exit after finishing their job
Environment=PORT=32768: sets the environment for the script (allows a bit of configuration)
After reloading the daemon to find our new service, we can run it with the following:
If we included an exit 1 in the script, the status of the service would be
failed even though it is oneshot and the log file will let us know the data
that failed it or if there was a connection refused (printed to standard
error). This allows systemd to answer the question “what services are currently
in the failed state” and I’d imagine that one could create generic alerts off
One of the last things we need to do is create a timer to be triggered every minute.
It might annoy some people that the amount of configuration is about the same number of lines as our script, but we gained a lot. In a previous version of the script, I was preserving standard out by using tee with process substitution to keep the script concise. This resulted in logs showing the script running every minute, but the data in graphite only captured approximately every other point. Since I knew from the logs that the command successfully exited, I realized process substitution happens asynchronously, so there was a race condition between tee finishing and sending the request. Simply removing tee for a temporary buffer proved effective enough for me, though there reportedly are ways of working around the race condition.
New script relies on system utilities and jq, which is found in the default ubuntu repo.
Logging output into journald provides a cost free debugging tool if things go astray
Anything other than what’s expected will cause the service to fail and notify systemd, which will try it again in a minute
Starts on computer boot
Sounds like an improvement!
Now that we have our data robustly inserted into graphite, now to time to graph it! The two data points we’re interested in are dns_queries_today and ads_blocked_today. Since they are counts that are reset after 24 hours, we’ll calculate the derivative so we can get a hitcount.
The best part might just be that I can add in a link in the graph that will direct me to the pi-hole admin in the situations when I need to see the full dashboard.
I wrote this article not because I’ve built a home server but because I’m on
the verge of doing it and I’d like to justify (to myself) why building one is
reasonable and why I chose the parts I did.
I want to build a home server because I realize that I have content (pictures,
home movies, etc). This content used to be stored on an external hard drive,
but when that hard drive died, I lost a good chunk of it. Since then, I’ve
moved the rest to a Windows Storage Pool. But then I thought about accessing
the content remotely, and I didn’t want my work / gaming PC to be on 24/7 and
exposed to the internet for power efficiency and security respectively. Having
an overclocked CPU and GPU on 24/7 (even if at idle) isn’t ideal.
Using Onedrive has fine – great for editing online documents, but space is
limited and I want a better sharing story (eg. loved one’s backup their
pictures here too). Though, probably the most important reason is because
setting up a home system sounds fun to me and it’s a good learning opportunity.
The case really defines the rest of the build, so I’m starting here. A small
form factor (SFF) case will limit oneself to more expensive components while a
larger case will take up more room. I waffled between many cases – I was
trying to get a small case that would fit on a shelf in the utility closet, but
wouldn’t compromise on the number of 3.5” drives. The height restriction meant
a lot of decent mini tower cases were excluded because even they were too tall.
Here were the contenders:
A SFF case with 8 hot-swappable 3.5” drive bays inside + more is quite an
achievement and is the only option when SFF is needed with an absolute maximum
number of drives. The downsides are that at $150 it was on the pricey side and
many reviews stated that thermal management was a challenge so aftermarket fans
with case modding is a necessity. This guy wrote an
solely to convince people not to buy the DS380B. Anyways, one goal of this
build is to keep cost and effort to a minimum, so this case was eliminated.
An HTPC that has a lot going for it. The horizontal design makes it a alluring,
as it can be placed on one of my cabinets. But with only four slots for 3.5”,
it would be limited as far as a storage server is concerned. Double parity RAID
would mean that half of the drives are redundant. The worst thing that could
happen would be running out of room and being forced to decide on whether to
buy bigger drives or get a dedicated NAS case.
Fractal Design Node 304
A SFF cube case that has 6 3.5” bays, goes on sale for $60, has great reviews,
and touted for the silent fans!? Sold.
Lian Li PC-Q25
Special mention must be made to Lian Li’s case which houses 7 drive bays, costs
more, and some (not many) have reported thermal issues.
I’ve decided on the Pentium G4600.
With Kaby Lake, pentium processors are blessed with hyper threading so their 2 physical cores become 4 logical cores.
All “Core” chips don’t support ECC memory (thus excluded)
Paying 15% more a 100 MHz boost made me exclude the G4620
I actually wanted top of the line integrated graphics card (Intel HD Graphics 630) because there won’t be a dedicated GPU in this box and I will cry if I was GPU limited anywhere.
Cheap! I’m going to grab it when the price hits $80
The server will sit idle most of its life, so no need to get a powerful CPU. In the future, if it turns out I need more horsepower, by then there should be a nice array of secondhand kaby lake xeons out there.
A Mini-ITX motherboard that supports ECC memory, socket 1151, and the Kaby Lake
basically makes our decision for us!
There were a couple of ASRock boards and I went with E3C236D2I, the one with
six SATA ports (same as case) with an added bonus of IPMI.
Unfortunately, a $240 price tag is a bit hard to swallow. There is definitely a
price to pay for keeping the size down, but using enterprise RAM!
Speaking of the RAM, I went with a single stick of 16GB ECC RAM. This may seem
odd, but I’ll try and explain. I’m using ECC memory because I want to be safe
rather than sorry and I’m not scrounging around looking for pennies so I can
afford it. I’m only interested in a single stick because buying 32GB upfront seems overkill,
I’m not made of money. Since the motherboard only has two DIMM slots I
wanted something significant that should last in the meantime.
On a side note, RAM is expensive right now, the 16GB of RAM is retailing for
$150 whereas it debuted at $75. Don’t worry, I have price triggers.
Even though the case supports tower CPU coolers and the Pentium G4600 comes with a stock cooler, I’ve opted for the slim aftermarket Notctua CPU Cooler: Noctua NH-L9i. The Noctua promises to be much quieter than the stock cooler. Since the fan is so slim, if I decide to get an even tinier case in the future, the fan will fit!
Since I won’t be overclocking the CPU, I’ll be able to use the low noise adaptor to make the cooler even quieter.
I went with the SilverStone 300W power supply.
I couldn’t buy anything that was below 300W (I was shooting for something around 200W). The reason is that power supplies were made to operate between 20% and 100% of their rated wattage. If I had gone with a 450W power supply (the next power supply in SilverStone’s lineup), I’d need a idle usage of at least 90W instead of 60W to get that guaranteed efficiency. Basically, this is me being environmentally conscious.
80+ bronze rating was distinguishing in this low of power range
SFX form will allow me to get an even smaller case in the future if needed
Is semi fanless (quiet). People report that only under extreme duress does the fan turn on
I already have a couple of 4TB Seagate 3.5” drives, so getting more of them is a logical choice. Ideally, I wouldn’t have to buy all of them up front, but that is cost of ZFS. Here’s to hoping a get a good deal on them!
One of the things I’m still pondering is what I should do about a bootable drive. I could drop down to using a RAID of 5 drives and get a different drive for the OS. Brian Moses uses a flash drive. I’m actually thinking of using my one PCIe slot to host a M.2 PCIe adapter and grabbing a Samsung 960 EVO or something similar. PCPartPicker doesn’t list the motherboard as capable of using the M.2, but we’ll see about that, as the motherboard manual specifically calls out instructions for M.2 NVMe drives.
Update: The motherboard does support M.2 drives, but only of the smallest kind (form factor 2242), which really eliminates potential drives. I’ll have to look towards 2.5” Sata SSDs.
After trying for a week to get FreeBSD and Plex working together, I gave up and have decided that Ubuntu 16.04 with docker is the way forward. Let me explain:
The first task was to determine whether to use a hardware RAID controller or a
software based one. Searching around, it became clear that a software RAID was
better due to costs and features from file systems like ZFS. Speaking of ZFS,
it’s the best file system for a home server, as it is built for turning several
disks into a one and features compression, encryption, etc.
Choosing ZFS, it would make sense if the OS was FreeBSD. ZFS and FreeBSD go
together like bread and butter. They are tried and tested together. Since I was
(and still am) unfamiliar with FreeBSD, I spent a week learning about jails and
other administrative functions. The concept of jails (application isolation
without performance cost) sounded amazing. Not to mention FreeBSD seemed like a
lightweight OS. Running top would only show a dozen or so processes. I got
quickly to work setting up a FreeBSD playground inside a virtual machine.
First I tried setting up an NFS server but ran into problems as I needed NFS v4
to run an NFS server on nested ZFS filesystems, but NFS v4 isn’t baked into
Windows, so it was a no go. Then after only a couple hours of fighting with SMB, I finally got it working. I’m just going to squirrel away the config here for a rainy day:
I think the trick was that I wanted SMB users to be users on the VM, so the
Samba server should act as the master.
So as you can see, everything was going smoothly – that is until I tried setting up Plex. I thought that since plexmediaserver was on FreshPorts that everything should work. It didn’t, and since I didn’t know FreeBSD, ZFS, or plex, I went on a wild goose chase of frustration. The internet even failed me, as the errors I was searching came back with zero results.
In a fit, I created an Ubuntu VM and ran the plex docker
container and everything just
worked. I gave up FreeBSD right then and there. I wasn’t going to force
something. I later found out that since FreeBSD represented less than 1% of
plex’s user base, the team didn’t want to spend the resources for updates. Oh well, ideally I wouldn’t have to use docker (downloading all those images seem … bloated), but since it’s rise to ubiquity and promise of compatibility, I’ll hop on the bandwagon.
With that, let’s take a look at some of the applications I’m looking to run:
ddclient: A dynamic dns client. It keeps my dns records updated whenever my ISP decides to give me a new IP.
nginx: A webserver that will serve as a reverse proxy for all downstream applications. Will be able to use certificates from Let’s Encrypt without configuring each application.
collectd: A system metric gather (CPU, Memory, Disk, Networking, Thermals, IPMI, etc). This will send the data to:
graphite: Using the official graphite docker image to store various metrics about the system and other applications. These metrics will be visualized using:
rstudio: The rocker docker image will let me access my rstudio sessions when I’m away. Currently, I have a digitalocean machine with rsutdio, but it’s been a pain for me to create and destroy the machine every time I need it.
pi-hole: blocks ads on the DNS level so one can block ads on all devices on the network. And, of course, there is a docker image, which has been working wonderfully in my test playground.
You’d be wrong if you thought I’d abandon my current cloud storage providers (Onedrive, Google Drive, etc). In fact, I pay them for increased storage because stuff happens and I need to have backups of pictures, home videos, code, and important documents. I’m planning on keeping all the clouds in sync with everything encrypted using rclone. That way if a backup is compromised, it is no big deal.
I’m also not going to abandon DigitalOcean, as those machines easily have more uptime and uplink than Comcast here. My philosophy is that if I want to show people my creation, I’ll host it externally, else I’ll self-host it. Plus it is a lot easier to tear down and recreate machines with an IAAS rather than bare metal.
The only question now is … when will I jump head first?