Ok so I've noticed for some reason it's not working anymore - I started it to solve recaptcha v2, v3 etc. but the command prompt just sits there in perpetuity without recognizing the queries...
I indeed updated Python, FFMPEG etc. but still the same problem... Port 8181 still open too...
daviddig1 San Antonio, TX
Ok so I think I solved my own problem: all I had to do was firewall inbound and outbound rule python's exe files as 'allow'
Botmaster / Xevil are mentioning Google ReCaptcha changes since Thursday this week and thus many failures.
daviddig1 San Antonio, TX
Thanks @organiccastle yes and even better than before as far as I can tell - after whitelisting python's exe files the number of positive solves seem to have gone way up
Standby as I test further...
daviddig1 San Antonio, TX
Yeah @organiccastle still so far so good it seems - I strongly recommend trying that potential solution yourself to see if that boosts your number of positive results
Thanked by 1organiccastle
daviddig1 San Antonio, TX
edited May 2024
Just one thing I wanted to mention: web contact does find false positives for recaptcha v2/v3, but that's only because of the sites' malformed scripting, NOT the fault of the program or the capv2v3 solver.
Therefore, I added the following code to trace those false positives as well as the more specific failures (added right after line 1255 in caphv2v3.py):
elif line.startswith('ERROR_RECAPTCHAV2_AUDIO_NOT_FOUND=The reCAPTCHA was not found'):
return PlainTextResponse('ERROR_RECAPTCHAV2_AUDIO_NOT_FOUND=The reCAPTCHA was not found')
elif line.startswith('EXCEPTION=Event loop stopped before Future completed.'):
return PlainTextResponse('EXCEPTION=Event loop stopped before Future completed.')
elif line.startswith('ERROR_RECAPTCHAV2_AUDIO_RATE_LIMIT=The reCAPTCHA rate limit has been exceeded.'):
return PlainTextResponse('ERROR_RECAPTCHAV2_AUDIO_RATE_LIMIT=The reCAPTCHA rate limit has been exceeded.')
Ok so I did also notice another boost in positive solves simply by adding a 5-second delay between page load and submission in web contact, since some sites take a while to load recaptcha after the rest of the page loaded (perhaps caused by 'defer' or 'async' [?])
daviddig1 San Antonio, TX
edited May 2024
TWO BIRDS WITH ONE STONE!
Just now I firewall outbound and inbound allowed ffmpeg's exe files, and the solving is even faster!
I was also able to increase the number of threads without causing throttling
Sven www.GSA-Online.de
original post was updated with v66 (v2<->v3 solve when failed, update of ffmpeg and Headless browser)
daviddig1 San Antonio, TX
edited May 2024
@sven you're the man! The process is faster, more accurate and lighter-weight!
Then again I just got done whitelisting ms-playwright's exe files, as I have a hunch the firewall could've been causing an issue there too
@sven about this new version: any chance you can tell the script to keep trying to solve a recaptcha with the error 'ERROR_RECAPTCHAV2_AUDIO_RATE_LIMIT=The reCAPTCHA rate limit has been exceeded' ? If so do you know how?
Sven www.GSA-Online.de
@CL67F is the one who is coding this. All fame belongs to him! Im just the messenger adding new updates here.
Hi @CL67F about this new version: any chance you can tell the script to keep trying to solve a recaptcha with the error 'ERROR_RECAPTCHAV2_AUDIO_RATE_LIMIT=The reCAPTCHA rate limit has been exceeded' ? If so do you know how?
daviddig1 San Antonio, TX
Also @CL67F : as much as python is RAM-heavy, a lot of the blame for CPU usage I can actually put on Windows 11's very own security service, which works overtime when I'm running the recaptcha script
While I have whitelisted the files and folders in question, it's still doing that and thereby unnecessarily pulling resources - any chance you can help me mitigate that issue?
daviddig1 San Antonio, TX
Ok so luckily I made it drastically drop by going into "Windows Security" -> "Virus and Threat Protection" -> "Manage Settings" -> "Exclusions" and adding the following folders (which applies to subfolders as well): C:\Users\[YOUR USERNAME]\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0 C:\Users\[YOUR USERNAME]\AppData\Local\Programs\Python\
I think I know what was going on: when I searched this same exact problem on Google, someone pointed out on the same issue to whitelist any cache folders - by whitelisting the above folders, which include caches, the CPU usage on Windows Security plummeted from 20% to 6%
CL67F Strasbourg (France)
edited May 2024
ERROR_RECAPTCHAV2_AUDIO_RATE_LIMIT=The reCAPTCHA rate limit has been exceeded #### Too much access from same IP #### #### You get this problem if you are not using enough Proxies ####
Thanked by 1daviddig1
CL67F Strasbourg (France)
edited May 2024
I have a new look at leak memory.
Python Uvicorn is known for having memory leaks. Uvicorn reaches easily 800MB RAM. #### The only solution at this time is to set properly the number of workers that depends on the ressources of your CPU ! ####
daviddig1 San Antonio, TX
edited May 2024
Thanks @CL67F thanks for all your work I'm purchasing extra ram (up from 16GB all the way up to 64) and am currently using 500 webshare proxies
ERROR_RECAPTCHAV2_AUDIO_RATE_LIMIT=The reCAPTCHA rate limit has been exceeded #### Too much access from same IP #### #### You get this problem if you are not using enough Proxies ####
Ok can you tell me how to timeout proxies when that happens for, say, 5minutes?
daviddig1 San Antonio, TX
Yep that jump from 16 to 64GB RAM really makes things easier - processing 2,000 at 30 threads now only takes 2hrs instead of 5-6 to accomplish
Yeah lol I did run v2 solves over the thousands the day before - I am thinking of buying 500 more webshare proxies
CL67F Strasbourg (France)
edited May 2024
ERROR_RECAPTCHAV2_AUDIO_RATE_LIMIT : 739
### Should be a proxy problem ... ### ### The number of solves for a proxy is a parameter set by Google. ###
daviddig1 San Antonio, TX
Thanks CL67F I did buy 500 more proxies and I'll let you know later how that goes
daviddig1 San Antonio, TX
Anybody have any luck using storm proxies? I'm very curious about it's effectiveness
daviddig1 San Antonio, TX
Ok so I finally got rid of windows defender hogging the memory once and for all: I simply had to turn off "real-time protection" while running the recaptcha solver
daviddig1 San Antonio, TX
Ok unfortunately (at least as far as I can tell) I'm not able to use storm proxies since the IP's given are gateways and are treated like direct connections on the recaptcha solver's end
Comments
Is this redundant if we have xevil going?
I indeed updated Python, FFMPEG etc. but still the same problem...
Port 8181 still open too...
Botmaster / Xevil are mentioning Google ReCaptcha changes since Thursday this week and thus many failures.
Therefore, I added the following code to trace those false positives as well as the more specific failures (added right after line 1255 in caphv2v3.py):
elif line.startswith('ERROR_RECAPTCHAV2_AUDIO_NOT_FOUND=The reCAPTCHA was not found'):
Just now I firewall outbound and inbound allowed ffmpeg's exe files, and the solving is even faster!
I was also able to increase the number of threads without causing throttling
@organiccastle did you have better luck like I have?
Hi @CL67F about this new version: any chance you can tell the script to keep trying to solve a recaptcha with the error 'ERROR_RECAPTCHAV2_AUDIO_RATE_LIMIT=The reCAPTCHA rate limit has been exceeded' ? If so do you know how?
While I have whitelisted the files and folders in question, it's still doing that and thereby unnecessarily pulling resources - any chance you can help me mitigate that issue?
C:\Users\[YOUR USERNAME]\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0
C:\Users\[YOUR USERNAME]\AppData\Local\Programs\Python\
I think I know what was going on: when I searched this same exact problem on Google, someone pointed out on the same issue to whitelist any cache folders - by whitelisting the above folders, which include caches, the CPU usage on Windows Security plummeted from 20% to 6%
#### Too much access from same IP ####
#### You get this problem if you are not using enough Proxies ####
#### The only solution at this time is to set properly the number of workers that depends on the ressources of your CPU ! ####
##################################
### The number of solves for a proxy is a parameter set by Google. ###
However, I do like it's versatility in scrapebox