mmkhitaryan

Disk space issue

In my case I used container which had size of 12GB. It contained yolo model for inference, so the container was large. By default the size of template for ECS ec2 is 30GB, so it could not conain entire OS+Image.

The solution was changing the template root storage to 100GB.

Card type issue

Another case of PROVISIONING stuck was with spot instances. I wanted to save on inference costs, so wanted to run spot instead of on-Demand.

I set the filter to GPU accelerator, and selected all the instances which had GPU. The issue was that there are also non-ndidia GPUs in AWS. So Nvidia driver linux got installed on AMD machines, and the tasks got stuck in PROVISIONING state.

mmkhitaryan

import datetime

from tortoise import Tortoise, fields, timezone
from tortoise.models import Model

NAIVE_DATETIME = datetime.datetime.strptime("2024-2-11","%Y-%m-%d")

class TortoiseModel(Model):
    name = fields.CharField(max_length=50, unique=True)
    last_register_date = fields.DatetimeField()

mmkhitaryan

function sleep(ms) {
  return new Promise(resolve => setTimeout(resolve, ms));
}

function generateRandomString() {
  const randomChar1 = String.fromCharCode(Math.floor(Math.random() * 26) + 65); // Random uppercase letter
  const randomChar2 = String.fromCharCode(Math.floor(Math.random() * 26) + 65); // Another random uppercase letter

  // Concatenate the two characters
  const randomString = randomChar1 + randomChar2;

mmkhitaryan

from collections import defaultdict

pid_to_command_line = {}

list_of_all_processes = []

import csv

field_names = None

mmkhitaryan

import r2pipe

r = r2pipe.open()

r.cmd('ood')
r.cmd('dcu @main')

last_pointer = '1'
while True:
    poiner_from_stack = str(r.cmdj('pxqj 8 @rsp')[0])

mmkhitaryan

# r2 /bin/ls
# . ./run_in_r2.py
# 1
# 2
# ... 1000+

import r2pipe

r = r2pipe.open()

mmkhitaryan

# python stack_seeker.py some_file_in_cwd
import r2pipe
import sys

r = r2pipe.open('/bin/ls')

r.cmd('ood')
r.cmd('aaa')
file_name = sys.argv[1]

mmkhitaryan

#include <stdio.h>
#include <unistd.h>

void
f (int n)
{
  printf ("Number: %d\n", n);
}

int

mmkhitaryan

My router has been crashing when my PC boots, and starts working good after 5 minutes of restart. So I decided to dig into reasons of the crash.

I used tcpdump and made it start on system boot. It collected all the packets, and then used tcpdump replay to try to reproduce the crashes.

When I replayed all the packets, the router crashed as suspected. But I needed to understand what specific packets were the reason of the crashing.

So I started cutting the file in half, (basically binary search) and seeing if the crash happens on. I ended up with ~20 packets, and then choose those packets that are on the scapy script.

I wanted to continue the research of the reasons of the vulnerability, emulate the router firmware and try to crash it. But I did not find router's working firmware anywhere.

mmkhitaryan

from threading import Lock, Thread

accountone = Lock()
accounttwo = Lock()

def transfer(accountone, accounttwo):
    accountone.acquire()
    accounttwo.acquire()
    print("Transaction done")
    accountone.release()