Padding Oracle Attack doesn't just let you decrypt arbitrary messages: it lets you encrypt arbitrary data as well.
CT: IV | C1 | C2 | C3
PT: P1 | P2 | P3
P3 = D(C3) ^ C2
P2 = D(C2) ^ C1
P1 = D(C1) ^ IV
Calculate right to left:
- Take a random C3 and random IV, and run padding Oracle attack and retrieve D(C3)
- Now we have C3, we can get C2 as well from above equations. And we will have C2 | C3 which will decrypt to P3
- Now try to get C1, for that we need D(C2), but we have C2, run the same POA to get D(C2)
- Now after getting D(C2) we can retrieve C1.
- We have C1, but not D(C1) so run POA again to get D(C1).
- Once we have D(C1), we can retrieve IV as well.
- Now we have all the components of CT, we can form the complete CT as IV | C1 | C2 | C3
Complete code here: https://t.co/mLBXkGyk5B
@devops_inside Sure, with you on this journey. If you stumbled upon any networking concept do let me know. Post your learnings here and also the articles if you encounter any great one.
We all know IP to MAC address mapping and how mac address helps in addressing the packet to the destination.
But if we discuss layer 2 forwarding, there is another table: Forwarding Database, that keeps MAC address to bridge port mapping. It is Layer-2 forwarding state mapping.
It is helpful when you have containers on the same docker bridge. The docker bridge will forward the packets from one container to another based on layer 2 forwarding techniques and not layer 3. And this is the time when FDB: forwarding databases mapping is helpful.
Docker bridge does not need to know the IP address of the destination container to send it the packet but it needs to know: Mac address of the destination container + veth bridge it is connected with. So it examines the mac address mapping with veth.
Somehow I feels that large application like linux/nginx/redis can not be built completely via AI only.
Human code and review is important. If AI writes the complete code, then it might be possible that there is certain areas of performance improvements as they can be overlooked by AI.
I feel writing via AI is completely okay for small features or systems or a certain piece but over the time a big system can become a mess.
If a system involves docker container, host system processes and has firewall rules involving ufw/iptables/nftables, what tools you use to determine the packet path from another machine??
How do you know which rule it is hitting and which it is avoiding ???
How do you inspect routing L2 state?? These below are the most important commands to inspect:
ip route get <ip_addr>
ip neigh show dev <interface>
bridge fdb show br <interface>
bridge link show
@AyushSarode07 This is one of the best ways to understand data structure, algorithms, patterns, best practices, big project maintenance, makefile and so on...
Currently tracing a packet execution through running linux kernel 5.4 and I am amazed how much utilities functions are there to help us track the function calls, arguments.
I see tracing is actually an skill in itself.
I don't understand one thing, why every target is projected to 2047.
Every politician is saying 2047. Why they are waiting for 2047, why their actions are not reflecting the changes happening.
They are just dividing the country based on casteism (UGC), looting the middle class as much as possible, enjoying their thrones, big cars and a convoy of 20 vehicles (no toll) with them.
If 2047 is their target, it means they will keep looting the country till 2047.
They have destroyed the country to the core and enough damaged has been done now.