Catching DeadlineExceededError

[gmalquestion]

A request in my app takes too long and it is terminated with a
DeadlineExceededError exception. I want to catch it and show some
civilized output if this happens.

The below code doesn't work. I uploaded the code and still nothing is
caught. The exception appears (debug is turned on) as before. What's
wrong with this approach?


from google.appengine.runtime import apiproxy_errors


...

def get(self):
try:
....
except apiproxy_errors.DeadlineExceededError:
self.response.out.write("hey")
return

...

[Roberto Saccon]

look at rietveld code base, one of the things I have learnt there is:


# DeadlineExceededError can live in two different places
try:
# When deployed
from google.appengine.runtime import DeadlineExceededError
except ImportError:
# In the development server
from google.appengine.runtime.apiproxy_errors import
DeadlineExceededError


hope that helps

Roberto

[gmalquestion]

On Aug 6, 7:06 pm, Roberto Saccon <rsac...@gmail.com> wrote:
> look at rietveld code base, one of the things I have learnt there is:
>
> # DeadlineExceededError can live in two different places

Thanks! It's working,

The name apiproxy should have tipped me off that it's for the dev
environment.

[Ozgur Cem S.]

I danced around the "DeadlineExceededError" error recently, and I was
never able to trap the DeadlineExceededError. try: catch
DeadlineExceededError : never made it to my level so I can go around
it. (maybe I was doing something wrong...)

So the question is, can you trap DeadlineExceededError in the
application level?

[Calvin Spealman]

Being able to catch it would defeat the purpose of it: stopping you
from taking up cpu time. Now, I do think that giving you enough
microseconds to make a civilized response to the user would be OK, but
after that an uncatchable makes sense.

--
Read my blog! I depend on your acceptance of my opinion! I am interesting!
http://ironfroggy-code.blogspot.com/

[Ozgur Cem S.]

Calvin, I agree with you. I came to the same conclusion after a while.
My problem was happening during a bulk/insertion of records into the
datastore. Therefore, the visitor-user experience was not a concern of
mine, since that particular process is admin-only, and admins can
wait ;). All I needed was to trap, ifI exceeded the deadline, so I
can hop on a "redirect" and proceed the process in a separate request.
Oh well, I ended up limiting myself with 30 or so insertions at a
time, and redirect, and redirect and redirect.... and keep redirecting
until I have nothing else to insert. :)

This method worked pretty good for me so far on http://newsxperiment.appspot.com
. During periodical updates, I need to insert about 400 entries in the
datastore. And without splitting into seperate requests, there seems
to be no way to get that many :) records in there at one shot.

Cem