42 lines
2.5 KiB
Markdown
42 lines
2.5 KiB
Markdown
|
As a privacy advocate myself I do my best to respect your privacy while simultaneously keeping my server safe. I also think it's
|
||
|
|
important to be transparent about this kind of stuff so I wrote this document to tell you how exactly I process or store your
|
||
|
|
information.
|
||
|
|
|
||
|
|
## DNS & SSL
|
||
|
|
Currently I use cloudflare's name servers, however cloudflare doesn't own my domain (I didn't purchase the domain from cloudflare)
|
||
|
|
nor it proxies any of the traffic. All of my records use the *DNS only*" mode, meaning they are just DNS records and they point
|
||
|
|
to my server, not cloudflare. This also means cloudflare doesn't control any of the SSL certificates. All the certificates are
|
||
|
|
stored on my server and they are created with Let's Encrypt. So I own the certificates and it's not possible for cloudlfare to
|
||
|
|
suddenly switch DNS records (it would break SSL).
|
||
|
|
|
||
|
|
## Usage metrics
|
||
|
|
I don't have any kind of request or traffic monitoring on my server. So no, your HTTP(S) requests or other network
|
||
|
|
connections are not processed to generate colorful graphs, pie charts and stuff.
|
||
|
|
|
||
|
|
You may have realized that in the bottom of this page there is total visitor number tracker. It is the only
|
||
|
|
usage/visitor metric tracking I have, which is implemented by website's API, which is free (as in freedom) so you can
|
||
|
|
go audit it's code yourself.
|
||
|
|
|
||
|
|
I want to make it clear that this metric tracker does not store any information about you or your HTTP(S) requests to
|
||
|
|
a database. It temporarily stores your IP address' SHA1 hash, in memory, so it doesn't recount the same visitor again
|
||
|
|
when they refresh the website or visit it multiple times in a short period of time. After a certain amount of requests,
|
||
|
|
your IP address' SHA1 hash will be removed from the memory and it will be replaced with a new visitor's SHA1 hash instead.
|
||
|
|
|
||
|
|
## Logs
|
||
|
|
All the HTTP(S) services are proxied with nginx, and nginx logs all of them to a file on the disk. This file (`access.log`)
|
||
|
|
is only readable by the root user, and it is contents are deleted every 4 hours (with `shred` to make sure it doesn't leave
|
||
|
|
anything on the disk). The logs include *only* the following information:
|
||
|
|
|
||
|
|
- Request time
|
||
|
|
- Requested host
|
||
|
|
- Requested path
|
||
|
|
- HTTP request method
|
||
|
|
- HTTP response code
|
||
|
|
|
||
|
|
This is the minimal information I need to trace any issues if something goes wrong, which is the main reason why I use
|
||
|
|
logging at all, to make it easier to find any issues.
|
||
|
|
|
||
|
|
## Data removal
|
||
|
|
If you want to remove any of your data from my server, [you can send me an email](mailto:ngn@ngn.tf). And yes this includes
|
||
|
|
removing usage metrics and logs.
|