• 43 Posts
  • 283 Comments
Joined 4 years ago
cake
Cake day: September 14th, 2021

help-circle

  • <rant>I just spent an absurd amount of time setting up my local windows environment to use node 10 and gulp with node-sass and node-gyp (complete BS). I also need to do the equivalent for our Ubuntu pipeline.

    And guess what, I just need to do the same for the new team member that just joined and messed his installation with recent versions of everything.

    Oh, still find that easy? Lets just throw in crowdstrike and netscope firewalls and proxies for added security and block 50% of node environment and break all TLS/SSL connexion because things are not difficult enough.

    Want to use git? Sure, it’s jsut 3x to 4x times slower than on Linux.

    Did not pay extra carefull attention or do not know the difference between CR, LF and CRLF? let me just make your git life miserable by changing the line endings on all your files. Good luck fixing it.</rant>



  • I have tried. It is very difficult. In fact tiktok and Instagram have so many domanin names and IP addresses that it is hard to list them all and block them.

    Blocking the main domain would prevent you from accessing the web site, but the mobile app would work. Then you can block the api and stuff, but clients already connected would still work because they rely on the CDNs. When you block a good bunch of the network, the app would just feel Brocken, not blocked. When users keep reloading they will find a CDN you have not blocked yet. Also, they will still get notifications because they are pushed via Google services. They won’t see them in the app however.

    So you just need to monitor the logs to keep blocking tiktok domains that pop up.

    If you can block via regex or joker chars (*) then, it would be easier for you.

    You can look up domains list of tiktok on the internet for a good start.

    For my experience, I just managed to make tiktok and Instagram broken enough to frustrate the users and discourage its use over the network.














  • I guess it is better to spend some little time cleaning up your existing install than doing a fresh one. From what you describe, you just have some leftover configuration files you can delete.

    Fresh installs is needed when you messed up your current install with broken packages or missing ones or missing configuration and you just don’t know how to clean it up.

    So as long as you can clean up your install, just clean it.



  • The only two important columns are “Local address: port” and “process”. The later is what process is listening whille the former is the interface that process is listening on and the port.

    So you see that I don’t have any process listening on any port other than 80 and 443 iin the host and the regular ones.

    That said, you containers will still listen on the ports you want but only on a virtual network interface.

    Basically you only need to publish ports 80 amd 443 on the container or pod you have your reverse proxy on. Other containers need to only be attached to the same network as you already did.