[Django]-Django UnreadablePostError: request data read error

Why this error is happening ?

because the server is recieving a malformed request, which can happen for many reasons. someone might’ve canceled loading the page, someone might have a crappy internet connection that cut out, cosmic rays could have flipped a bit.

it’s not something you really need to worry about until it starts happening very frequently. you might want to make a note when these errors happen and see if it’s consequently on the same page or not.

How to solve?

you can’t. not at this point at least. gather some more data on when this bug occurs exactly. see if you can find a way to trigger it manually.

With an application at scale, you’ll always get the occasional cancelled request. If you’re receiving 500 emails they can be quite tedious.

I wouldn’t advise completely ignoring them. If UnreadablePostErrors are pouring in, then something’s wrong e.g. longer response times causes users to cancel requests.

My solution was a custom filter for admin emails where you can put any logic you want.

The simplest is probably to randomly ignore 19 out of 20 UnreadablePostErrors. That way, if something goes wrong i’ll still be informed, but i’ll be pestered 20x less.

If you want something more fancy, I’d go @pztrick’s solution.

import logging
import random
from django.http import UnreadablePostError

class ReduceUnreadablePostErrors(logging.Filter):
    def filter(self, record):
        if record.exc_info:
            exc_value = record.exc_info[1]
            if isinstance(exc_value, UnreadablePostError):
                return random.randint(1,20) % 20==0
        return True

settings.py:

'filters': {
   'reduce_unreadable_post_errors' : {
        '()' : 'path.to.your.ReduceUnreadablePostErrors'
    },

    ...

'handlers': {
    'mail_admins': {
        'level': 'ERROR',
        'filters': ['require_debug_false','reduce_unreadable_post_errors'],
        'class': 'common.utils.log.AdminEmailHandlerWithEmail'
     },

     ...

I wrote a throttled version of this. You’ll want to see the error if it’s spiking, but if it’s coming in every five minutes or whatever, you can ignore it. This assumes that if it came in twice in a one-minute period, you may have some kind of issue to look into.

def skip_unreadable_post(record):
    if record.exc_info:
        exc_value = record.exc_info[1]
        if isinstance(exc_value, UnreadablePostError):
            cache_key = "settings.unreadable_post_error"
            r = make_redis_interface("CACHE")
            if r.get(cache_key) is not None:
                # We've seen this recently; let it through; hitting it a lot
                # might mean something.
                return True
            else:
                # Haven't seen this recently; cache it with a minute expiry,
                # and don't let it through.
                r.set(cache_key, "True", ex=60)
                return False
    return True

...
    "filters": {
        "skip_unreadable_posts": {
            "()": "django.utils.log.CallbackFilter",
            "callback": skip_unreadable_post,
        },
    },
...
        "django.server": {
            "level": "INFO",
            "class": "logging.StreamHandler",
            "formatter": "django.server",
            "filters": ["skip_unreadable_posts"],
        },