If you really must do SSL with the SDK you can nginx as a reverse proxy in
front of the GAE SDK and have nginx handle HTTPS and serving the
certificate.
On Tuesday, October 8, 2013 8:10:22 AM UTC-7, Chetan Dhembre wrote:
Hi,
I have created one appplication using google app engine sdk in
Might be a different problem but one of our Python apps just started (half
an hour ago) having very slow warmup requests that are frequently throwing
DeadlineExceededError in Google code and requests that succeed are often
taking 10 seconds. Seems that it is starting to get resolved now though.
I used the form and got a reply stating that they are aware and looking
into it.
I was able to push through manual payments (using the same payment method
that failed) and the reply from the form confirmed that those had gone
through.
- Bryce
On Thursday, August 15, 2013 4:35:28 PM UTC-7,
Mandrill works great as well.
On Wednesday, July 24, 2013 1:22:13 PM UTC-7, Joshua Smith wrote:
Dunno how I missed this, but given the train wreck of GAE's email
delivery, this is very promising:
https://developers.google.com/appengine/articles/sendgrid
-Joshua
--
You received
My blobstore filenames are definitely not unique. In many cases there
are over 50,000 files with the same default filename. I don't use the
filename for anything. I just use the blobkey. Also, my users upload
files and I have never had a need to check for filename uniqueness.
This could be an
Regarding the Files API deprecation:
Is this as of right now or as of the final release of SDK 1.8.1?
I guess files already written will continue to work as is since they are no
different than other blobstore files? Anything wrong with that assumption?
With this announcement, Google's push for
did the job I've never investigated it)
On Tuesday, June 4, 2013 11:45:35 PM UTC+3, Bryce Cutt wrote:
Regarding the Files API deprecation:
Is this as of right now or as of the final release of SDK 1.8.1?
I guess files already written will continue to work as is since they are
no different
+1 for adding an optional daily limit for premier accounts. Obviously I
work to remove this sort of issue but bugs happen and I don't see an easy
way for me to see my monthly bill amount before I am actually billed. I
kind of miss daily usage reports with dollar values
On Tuesday, May 28,
No problem.
On May 24, 2013, at 6:12 AM, Joshua Smith joshuaesm...@charter.net wrote:
Yes! Thanks so much!
On May 23, 2013, at 6:22 PM, Bryce Cutt pandas...@gmail.com wrote:
Does this work for you?
handlers:
- url: /_ereporter.*
script
Had the same issue. Seemed to subside after about an hour and then was
performing very slowly for a while longer. Seems fine now.
On Friday, May 24, 2013 7:12:57 AM UTC-7, Iron Mountain Foundry wrote:
Good news, it was fixed overnight. No reply from Google on the cause or
the fix.
Mostly
Does this work for you?
handlers:
- url: /_ereporter.*
script: google.appengine.ext.ereporter.report_generator.application
login: admin
On Thursday, May 23, 2013 6:32:17 AM UTC-7, Joshua Smith wrote:
Is it possible to use ereporter in a python 2.7, threadsafe application?
I can't find
I agree with Vinny. I managed to make things a bit better in the past by
getting DKIM and SPF working for my Google Apps domain but still had lots
of email blocked with no reporting telling me what happened.
I've tried a few third party email solutions and so far my favorite is
Mandrill. I
I use Aptana Studio (which is just a dressed up version of Eclipse)
primarily and Sublime Text sometimes. Eclipse has been such a core part of
my workflow for over a decade that I always miss things about it when I try
to use other IDEs. I've been trying to switch away from it for a while but
AFAICT: He is OK if he is not doing this to avoid fees and the apps do not
work together as a single app. But I'm no lawyer so don't take my word for
it.
On Monday, April 22, 2013 1:24:18 PM UTC-7, barryhunter wrote:
Sounds like 4.3(e) would be a problem here.
If you want your query results ordered by flavourID then specify that as
the property in the order by clause of your query:
SELECT * FROM flavours ORDER BY flavourID
I suggest you read up on App Engine keys and ids:
https://developers.google.com/appengine/docs/python/datastore/entities
On
GAE supports outbound sockets with SSL as of SDK 1.7.7 but it is still
experimental.
https://developers.google.com/appengine/docs/python/sockets/overview
You need to include SSL in your app.yaml and it is only supported by the
Python 2.7 runtime.
djangoforms ModelForm is the same as the regular Django form class
(django.forms.Form) but with additions that allow it to interact with
datastore models.
If you want to keep most of your form code and only have to update any of
the GAE model specific stuff you do you can just include the
Jason,
Have you thought of using transactional tasks to update the search index?
Or are you more concerned with concurrent updates to the same model?
- Bryce
On Wednesday, April 17, 2013 3:08:05 PM UTC-7, Jason Collins wrote:
We've been doing a fair amount of work with Search API lately.
.
Arie.
On Tuesday, April 2, 2013 12:04:56 PM UTC-7, Bryce Cutt wrote:
IIRC it does not. Works fine on blobstore though.
On Monday, April 1, 2013 4:46:47 PM UTC-7, Jason Collins wrote:
We use the Datastore Admin backup tool (fired by a cron) to backup our
data into Cloud Storage nightly. Our
IIRC it does not. Works fine on blobstore though.
On Monday, April 1, 2013 4:46:47 PM UTC-7, Jason Collins wrote:
We use the Datastore Admin backup tool (fired by a cron) to backup our
data into Cloud Storage nightly. Our storage costs have crept up and I'd
like to delete the old backups.
Debugging is an absolute necessity. I use Eclipse PyDev (well, actually
Aptana Studio) specifically because of the graphical debugger. I feel blind
when I don't have a debugger; It's one of the most useful tools I know of
for development. I'm sticking to 1.7.5 for this (and other) reasons.
-
I have noticed this as well. Going into the instances control panel for the
old version and manually stopping them fixes it for me but it can take a
bit of time for new version instances to be used.
Also, I have become very used to doing everything I can to not break the
schema across versions
If you want an older version just copy the download URL for the current
version, change the version number in the URL to the one you want, then
open that URL. They keep a lot of older versions available for download.
On Wednesday, March 20, 2013 4:04:03 PM UTC-7, jay wrote:
Also, quick
A lot of this can be done in Python and since you need this in
functionality in Java you can set up a Python version or backend wrapped in
a RESTful API that uses xhtml2pdf and then consume that API from your Java
code.
For HTML/image to PDF I use the Python library http://www.xhtml2pdf.com/
I have heard good things about Mailgun but have not used it myself. I have
used sendgrid.com and mandrill.com and both work well. I like the
flexibility of Mandrill being able to send from multiple domains with DKIM
and SPF but the whitelabelling in Sendgrid is more complete. 12,000 free
Not sure if it is the same issue but my symptoms are similar. After
updating to SDK 1.7.3 on Windows 7 I have noticed that the compiled Python
files (.pyc files) are not always overwritten when I modify a source (.py)
file and it is executed. I never noticed this before SDK 1.7.3 and I have
First read the first few bytes of the file using a BlobReader and check to
make sure they contain a magic number for an image format you support.
This quickly rules out non-image file types. If this test fails then
discard the image.
Then load the image data and create an Image object based on
That sounds pretty consistent with my own experience.
https://developers.google.com/appengine/docs/python/memcache/overview#How_Cached_Data_Expires
If you want better guarantees you may want to star this issue:
http://code.google.com/p/googleappengine/issues/detail?id=6078
On Tuesday,
Correction: The necessary argument is --address=0.0.0.0
On Monday, September 24, 2012 12:20:14 PM UTC-7, Bryce Cutt wrote:
I have seen this before and I believe it is due to the browser making
multiple parallel connections to the dev server. The dev server only
supports a single connection
Waleed, VIP does not solve this issue. You get the same SSL cert warning.
They imply that the IP could change at any time so I would not rely on it
being the same for multiple requests.
Thomas, If you just want a redirect you could set up a VPS that does the
redirect for you and uses either
I have seen this before and I believe it is due to the browser making
multiple parallel connections to the dev server. The dev server only
supports a single connection at once so some/many of these connections will
time out or just fail. It is worse if you have multiple browsers making
http://code.google.com/p/googleappengine/downloads/list
A ton of bug fixes in this one.
App Engine Python SDK - Release Notes
Version 1.7.2
===
- Paid applications can now upload Static files and Code 1GB. Additional
storage will be billed at $0.13/GB per month.
:20 PM UTC+3, Bryce Cutt wrote:
It is in the request headers. See X-AppEngine-Country,
X-AppEngine-Region, X-AppEngine-City, and X-AppEngine-CityLatLong:
https://developers.google.com/appengine/docs/python/runtime#Request_Headers
https://developers.google.com/appengine/docs/java/runtime
It is in the request headers. See X-AppEngine-Country, X-AppEngine-Region,
X-AppEngine-City, and X-AppEngine-CityLatLong:
https://developers.google.com/appengine/docs/python/runtime#Request_Headers
https://developers.google.com/appengine/docs/java/runtime#Request_Headers
On Tuesday, September
One of my Python apps that usually had between 6 to 8 second long warmup
requests now seems to have 1.5 second warmup request today.
I wondered if perhaps the app was just moved to a less noisy cluster so I
tested 4 other Python apps I have and confirmed that all of them now take
around 1 to
.
I'd love to hear about a better solution though...
-Robert Fischer
www.DealScorcher.com
On Tue, Aug 28, 2012 at 10:31 PM, Bryce Cutt pand...@gmail.comjavascript:
wrote:
I am having an odd issue with Python 2.7 and SQLite and am wondering if
anyone has seen it or has some insight.
I
I am having an odd issue with Python 2.7 and SQLite and am wondering if
anyone has seen it or has some insight.
I recently decided to try porting some of my Python apps to the python27
runtime and the first step was to install Python 2.7 on my dev machine
(Windows 7 64bit). Now whenever I run
serially but in production they are processed
simultaneously.
I'm curious how you plan to benchmark the raspberry pi dev_appserver vs
the production stack.
-Robert Fischer
www.DealScorcher.com
On Sat, Aug 25, 2012 at 10:05 PM, Bryce Cutt pand...@gmail.comjavascript:
wrote:
Some
I suspect you are very right Anand. It seems likely that the speed of the
RAM is going to play a huge part as well. I didn't mean to suggest it was
an apples to apples comparison, just proposed it as a curiosity.
If I was going to use the Pi as a web server so that it could perform a
task in
Some of you are probably aware of the Raspberry Pi single board
computerhttp://www.raspberrypi.org/faqs that
runs Linux and Python quite well. When I first heard of it I had lots of
ideas of what to use it for and one of those was to run App Engine on it.
Why? Because I can. :)
My Raspberry
Other than the one outlier at the top of your image your numbers look
pretty typical to me.
Memcache can have a very variable response time. I have seen them go as
high as 1 second on a bad day but this has been rare. On a usual day I see
average times around 5ms to 20ms with occasional spikes
Related to discussion on Twitter Dev board:
https://dev.twitter.com/discussions/3218
On Friday, August 24, 2012 7:26:34 AM UTC-7, JH wrote:
As of yesterday my app, ticker-app, can no longer connect to Twitter's
API. It has worked fine for over a year. As of yesterday any urlfetch
call to
the release in the announcement blog
posthttp://googleappengine.blogspot.com/2012/08/app-engine-171-released.html
.
Thanks,
Dan
On Tue, Aug 21, 2012 at 11:55 AM, Bryce Cutt pand...@gmail.com wrote:
Release notes are up too:
http://code.google.com/p/googleappengine/wiki/SdkReleaseNotes
Thanks.
On Wednesday, August 22, 2012 1:01:50 PM UTC-7, Jason Collins wrote:
Look at
google_appengine/google/appengine/ext/appstats/sample_appengine_config.py-
there's all sorts of good stuff in there.
j
On Wednesday, 22 August 2012 13:39:56 UTC-6, Bryce Cutt wrote:
This is great
http://code.google.com/p/googleappengine/downloads/list
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To view this discussion on the web visit
https://groups.google.com/d/msg/google-appengine/-/jAx8jUVb0mgJ.
To post to this group, send
Release notes are up too:
http://code.google.com/p/googleappengine/wiki/SdkReleaseNotes
Personally I am most interested in AppStats showing information about the
cost of the RPCs made during a request.
On Tuesday, August 21, 2012 11:35:10 AM UTC-7, Bryce Cutt wrote:
http://code.google.com/p
https://developers.google.com/appengine/docs/ssl
On Tuesday, June 26, 2012 3:54:54 PM UTC-7, James Gilliam wrote:
Looks like they are deploying version 1.7 with SSL stuff ...
There is no blog posts describing the function ... SSL VIP $100 a month
... pricey
--
You received this message
http://googleappengine.blogspot.ca/2012/06/google-app-engine-170-released-at.html
On Tuesday, June 26, 2012 3:54:54 PM UTC-7, James Gilliam wrote:
Looks like they are deploying version 1.7 with SSL stuff ...
There is no blog posts describing the function ... SSL VIP $100 a month
...
If you have large amounts of text in a single field you should get better
compression from zlib than Snappy will get so your data will still be
smaller if you do your own compression. YMMV.
Also, if you have large text fields you may save on storage costs by
storing this data in the blobstore
aschmid: The ndb BlobProperty has optional compression built in (see
ndb.model.BlobProperty). You could implement the MarshalProperty like this:
class MarshalProperty(BlobProperty):
def _to_base_type(self, value):
return marshal.dumps(value, MARSHAL_VERSION)
def _from_base_type(self,
Hi Andrin,
I have run a few more tests and now I am convinced you are right.
On Python 2.5 where pickle and json are both implemented in pure Python the
speed benefits of marshal outweigh any conversion issues in the future
(which are not likely anyways). Your numbers show even on Python 2.7
Andreas,
Yup. I have had to resort to using the blobstore on many occasions for
exactly this reason.
One gotcha that I have run into when doing this is that there appears
to be no way to write a new blob to the blobstore (using the files
API) inside of a transaction that also modifies a
The docs for marshal seem to indicate there are no guarantees marshaled data is
compatible between python versions. That worries me. If I decide to eventually
upgrade my python 2.5 apps to 2.7 am I going to have to convert all my data
between marshal versions? While pickle is not always
The type of data being serialized certainly affects how much faster marshal
is. When testing with just a string approximately 1MB in size marshal was
around 10 times as fast as JSON but only about 10% faster than pickle. With
a dict of dicts of integers (around 1MB when serialized with marshal)
Just copy the feedparser.py file into your project. You can download the
full package here (https://code.google.com/p/feedparser/) but you only need
the one file.
#to use it
import feedparser
feed = feedparser.parse('some_url')
for entry in feed.entries:
print entry.get('title')
On Friday,
As Nick said on his blog:
http://blog.notdot.net/2010/01/ReferenceProperty-prefetching-in-App-Engine#comment-533928983
The SDK was modified a while ago to send one request per entity group in
production. The requests are sent asynchronously, and the new method is
actually faster than sending
http://code.google.com/p/googleappengine/issues/detail?id=7019
On Feb 29, 5:28 am, Brian Davenport the.lost.min...@gmail.com wrote:
Looking at the traffic splitting box has a getting started link to what I
assume is a internal resource at google
When starting the development server set the --datastore_path to a
path that is not in a temporary directory. You may also want to set
the --history_path option (but I have never had to).
http://code.google.com/appengine/docs/python/tools/devserver.html#Using_the_Datastore
On Feb 22, 8:24 pm,
A quick solution for you would be to switch to Searchify (http://
www.searchify.com). Searchify has taken the open source code from
Indextank and set up up a duplicate service. I have switched over one
of my Indextank sites to their service and things have been working
well so far. They mentioned
Brandon,
For some reason the web view of this group is not always recognizing
the quoted portion of your messages so it is not hiding them with a
nice Show quoted text link.
Also, I mostly read this group through the digest emails and the
digest generator seems to be smart enough to cut off the
:
This is actually a very good question; I'd like to know the answer as well.
Robert
On Mon, Jan 23, 2012 at 23:31, Bryce Cutt pandas...@gmail.com wrote:
My mistake. I really should RTFM.
Is the cost per thousand estimate based on the new pricing model or
the old one?
On Jan 23, 8:39 pm
, Google App Engine
plus.ikailan.com
On Fri, Jan 27, 2012 at 12:32 AM, Bryce Cutt pandas...@gmail.com wrote:
I would definitely be interested in a replacement that is based on the
new pricing.
Any suggestions on a good way to estimate costs using information that
is currently
I am wondering if the cost per million estimate (reported in logs as
cpm_usd) is based on the new pricing model or is it still based on the
old pricing model?
Does anyone who is getting lots of hits have any numbers that show how
accurate of an estimate cpm_usd is?
--
You received this message
I have run into this issue before.
The dev server only handles one request at a time. This problem could
have something to do with the browser making multiple requests to the
server at the same time (like Chrome often does) or it could be that
you used a different browser to connect to the server
to start the application I got a message from Windows
Firewall and I had to allow access to pythonw.exe...I never got that
message before, so I guess the --address=0.0.0.0 triggers it.
Thanks a lot.
On Jan 23, 7:30 pm, Bryce Cutt pandas...@gmail.com wrote:
I have run into this issue
My mistake. I really should RTFM.
Is the cost per thousand estimate based on the new pricing model or
the old one?
On Jan 23, 8:39 pm, Richard Watson richard.wat...@gmail.com wrote:
CPM = cost per mille, or thousand. That 1000-fold difference would cost
you far more than multiple pricing
things
differently or the bulk get is actually being called multiple times
when it only should be called once.
On Dec 3, 5:48 am, Timofey Koolin timo...@koolin.ru wrote:
Production get every entity too. But do it in parallel mode - at the same
time.
2011/12/3 Bryce Cutt pandas
Also posted on SO:
http://stackoverflow.com/q/8390052/98138
I have a model with 15 properties and indexed=False set for each of
them. When I write an entity of this model it does not write to the
property indexes for those properties however when I transfer that
model from one app to another it
Did the production server change the way it handles a db.get() on a
list of keys? I believe it used to do a single get_async on the list
of keys but now it appears that it is doing a get_async for each key
(according to appstats). The development server is still doing a
single get_async for the
A description of the steps (with examples) can be found in the answers
here:
http://stackoverflow.com/q/4994913/98138
As Niklas mentioned things will be a lot easier with the new Python
runtime but for now the easy fix is at the link I provided.
Some things changed between Django .96 and 1.2 and
Seems to be fixed today with the release of SDK 1.6.0. I just re-
enabled AppStats on the live site and am no longer seeing these errors
in the logs.
On Oct 6, 1:01 am, J.T gae.0...@gmail.com wrote:
E 2011-10-06 10:01:05.834
/base/python_runtime/python_lib/versions/1/google/appengine/ext/
I am seeing these too.
I have reported it as a production issue:
http://code.google.com/p/googleappengine/issues/detail?id=6056thanks=6056ts=1317920588
I'll also CC Johan as this seems to be affecting lots of us and he'll
know who to forward it to.
- Bryce
On Oct 6, 2:01 am, J.T
+1
On Oct 4, 1:58 pm, Francois Masurel f.masu...@gmail.com wrote:
Am I the only one ?
Francois
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To post to this group, send email to google-appengine@googlegroups.com.
To unsubscribe from
Working for me now.
On Oct 4, 2:22 pm, Bryce Cutt pandas...@gmail.com wrote:
+1
On Oct 4, 1:58 pm, Francois Masurel f.masu...@gmail.com wrote:
Am I the only one ?
Francois
--
You received this message because you are subscribed to the Google Groups
Google App Engine group
The edge cache is just that - a cache. It doesn't guarantee that no 304
requests will reach your app, only that they will be satisfied by the cache
if possible - and if the headers you set permit it. The more popular your
content is, the more effect you're likely to see from the cache.
The onus is always on you to have as many levels of cache in place as
you can get away with so, depending on what you are caching, you are
going to benefit from using the edge cache, memcache, and instance
global variable caching all at the same time.
Sure, but when designing our apps we
76 matches
Mail list logo