increase speed 50x with using set()
plaber opened this issue · 30 comments
if you change some code to cache list of wallets in ram
line = open('bit.txt').readlines();
line.sort()
lset = set(line)
while True:
i=i+1
..................skip some code.....................
#print ('Worker '+str(r)+':'+ str(i) + '.- # '+pub + ' # -------- # '+ priv+' # ')
pub = pub + '\n'
if pub in lset:
msg = "\nPublic: " + str(pub) + " ---- Private: " + str(priv) + "YEI"
text = msg
now it show 1200 K/s instead of 45 K/s for list of 1M wallets
Once modified with load in RAM (set) it gives me 540 K/s * core. Executed on a raspi 4 cores in total at 2160 K/s for list of 33M wallets
@Plaztomad I can share it but it is not enough to get to achieve something. In addition, there are other software that use gpus that are much more powerful and do not achieve anything due to the astronomical number of possible combinations. All of this is just for experimentation.
@Plaztomad Modified file attached. Try it and tell me how you see it.
seekanddestroy.zip
Everything has been tested on ubuntu server 20.04 with a raspberry 4 and it works perfectly.
Improvements made:
- database with 33M wallets
- speed increase with load in RAM (set) of the database
- send mail if there is success.
- 2160 K/s for list of 33M wallets
Hardware requirements:
- 8GB RAM minimum
- 4 cores minimum
the path to the database is indicated on line 22
The database I use contains about 33 million wallets. Download it from here: database
Lastly, the database surely contains both compressed and uncompressed addresses and this software only generates uncompressed addresses. It would be necessary to add that function.
working great! thanks for your support.
@Plaztomad Modified file attached. Try it and tell me how you see it. seekanddestroy.zip
the path to the database is indicated on line 22
The database I use contains about 33 million wallets. Download it from here: database
Lastly, the database surely contains both compressed and uncompressed addresses and this software only generates uncompressed addresses. It would be necessary to add that function.
I get error:
F:\BitBruteForce-Wallet-master>python seekanddestroy.py
reading database: 35/35 Traceback (most recent call last):
File "F:\BitBruteForce-Wallet-master\seekanddestroy.py", line 133, in
database[3] = database[3] | set(pickle.load(file))
_pickle.UnpicklingError: could not find MARK
@ImpAsh13 I run it on ubuntu 20.04 and it works perfectly.
does it work with the original code?
have you installed the requirements?
Do you use python3?
Did you download the database in the correct path?
database/MAR_15_2021
@Plaztomad Yes, but you have to pass it to pickle format
I run it on ubuntu 20.04 and it works perfectly.
does it work with the original code?
have you installed the requirements?
Do you use python3?
Did you download the database in the correct path?database/MAR_15_2021
I run it on Windows 10 Pro
With original code it work.
All recommendations downloaded and installed, except for 3 billion years.
Python 3.10
Database is loaded. All 34 parts. After that, 10 seconds pass and an error message appears.
I will try to run on ubuntu
I found a bug on my computer and was able to reproduce it.
It was all about the disabled paging file.
@ImpAsh13 What type of processor and how many cores does it have?
Loading the entire database into memory requires at least 8GB of RAM
I don't think it will work with less than 8GB RAM
@jomagalo
I have Intel(R) Core(TM) i7-7700 CPU @ 3.60GHz 8 cores and 16 Gb RAM.
Swap file need for pickle.
Now speed of each core ~1200 K/s
@ImpAsh13 It may be the version of python you have installed, I use Python 3.8.10 and everything works fine.
Check also this part that I have updated: #28 (comment)
wow 1200 K/s * core
@jomagalo
Here #28 (comment) I wrote that everything worked for me after I enabled the swap file.
ok, I had not seen it
eye in the email sending configuration, you must put your data. Don't use gmail seems to give problems. With any other mail server it works fine.
@Plaztomad Yes, but you have to pass it to pickle format
Please any hint or guideline to do that?
@Plaztomad Sorry, I have no info on that. I took the database as it is.
I update base, but how to check it?))
@jomagalo
What printing me:
Worker 0:27654.- # 1GZZ95CPME4hpqkxUXvH7ws5sYVcRTWRPm # -------- # 5JTsQHSr2UF9ZYmmoRgNqGeUfTKYXEtYN5keLcEYMMkeZjkWcv7 #
rand address and key
thats normal?)
@ImpAsh13 I don't know what kind of check you want to do. If what you want is to verify that the content of the database in memory works just uncomment lines 65, 66 and 67. What this does is verify after 5k random checks check this wallet 1L1aAhBakQGrhzdJG4iJMp77c7TqVk5bWj with the database in memory and since it obviously exists, it will show you the data.
code line 69 what it does is show the internal operation, that is, it prints each of the random wallets it generates.
@jomagalo
I want to check the connected database with addresses in pickle-file.
Did it really load in the right format, because I converted it from csv file to pickle.
Is it possible to display a single line or entry from a given pickle-file?
@ImpAsh13 You can do what I tell you in the previous post. Take a wallet that you know is included in the pickle file and paste it on line 67 of the code. You must uncomment lines 65, 66 and 67
@Plaztomad I can share it but it is not enough to get to achieve something. In addition, there are other software that use gpus that are much more powerful and do not achieve anything due to the astronomical number of possible combinations. All of this is just for experimentation.
Hi, I've got a lot of GPU, where can I find this software?Thank you
@jomagalo When I created my pickle file and replaced those 34 from the database folder I get an error:
reading database: 1/2 Traceback (most recent call last):
File "/seekanddestroy.py", line 102, in
database[1] = database[1] | set(pickle.load(file))
_pickle.UnpicklingError: invalid load key, '\x00'.
Can you tell me how to fix it?
my code for 1000 K/s