Hi Lorne
>
> Here is the link https://issuetracker.google.com/issues/65781183
>
> The DNS changed in cloud console on reverification, and our site is
> accessible using the name.
>
> So I don't have any evidence to suggest it didn't .
>
> Regards
>
> Tim Hoffma
Hi Lorne
Here is the link https://issuetracker.google.com/issues/65781183
The DNS changed in cloud console on reverification, and our site is
accessible using the name.
So I don't have any evidence to suggest it didn't .
Regards
Tim Hoffman
On Tue, Sep 19, 2017 at 1:00 AM, '
HI
I store apostrophe's in the datastore all the time.
You should first check the strings as you get it out of Youtube.
Log them, or try intreractively through the interpreter without appengine
in the way, or use pdb, then
set breakpoints all the way to just before you put the record and see i
Hi
Its not generally considered an error if something is not in the cache, (it
is after all a cache, not a guarunteed storage.)
Use a decorator to raise KeyError or whatever you feel is appropriate.
Rgds
Tim
--
You received this message because you are subscribed to the Google Groups
"Googl
Well I have 1000+ users with an app and every page is absolutely unique to
them and no anonymous users, and the contents of every page is dynamically
calculated and will be updated any time they do anything, they are required
to log in every week or so to log their work hours to comply with thei
Infinite and completely useless for some types of apps ;-)
T
--
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/-/jlGF0wQmxJcJ.
To post to this group,
Hi Brandon
memcache effectiveness is going to be heavily dependent on what you are
caching.
I have in the past had memcache up to 70MB (but I haven't looked at it in a
while.)
One of my sites has a 76% hit ratio on memcache, but it only uses <7MB and
the app has a small data set.
I should use
When you start the dev server use --default_paritition argument to override
the dev~ name used. So you would use
See http://code.google.com/appengine/docs/python/tools/devserver.html
T
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To
Hi Greg
Have a look at my profile, not too many posts lately.
http://groups.google.com/groups/profile?user=zutesmog%40gmail.com&qt_a=Look+up+author
I only post to a couple of groups.
You can see my reply to "Can I use set in appengine?" isn't coming up in
appengine but is listed in some wierd
This seems widespread I am finding some of my posts here appearing in adult
content groups.
--
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/-/s
Hi PK
Not sure what you call a lot of data. I have successfully imported 500,000
entities into the sqlite backend.
The trick I found was to specify single thread during the bulk upload. I
found with multiple threads the dev server
would ultimately lock up or the bulk upload would fall over wit
Actually I am stupid, I just realized that the actual charge is for the
minimum billing spend.
They are just inflating the used hours in the statement for some reason,
which through me off.
That shouldn't be necessary, they should just rely on the minimum spend
line item.
Also the billing
Actually on closer examination there is no period assigned to the bill and
I can assume its for the most recent billing event.
Serious lack of information going on here
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To view this discussio
Hi
I just recieved a bill (admittedly small) for the 2011-11-20. In my
billing in history I have in fact
two entries for that day, and in fact for the same time. One has billable
hours of 117 and the other 16.7
This has never happened before (that many hours used and two entries in the
billi
Static files (mapped by the static handler) are not accessible from your
application code.
The static files are handled by different infrastructure. You will need to
deploy a second copy
one for your application code access and one for static handler.
Its always been this way.
Use a symlin
Hi
If your using Python 2.5 then you need to import the sets module.
>>> from sets import Set
>>> Set([1,2,1,2,4])
Set([1, 2, 4])
>>>
This code will also work in 2.7
Though in 2.7 you have the set type builtin.
Rgds
Tim
--
You received this message because you are subscribed to the Google
HI Brian
You should put that fact in a prominent position in the 1.5.5 release notes.
Tim
--
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/-/KmKduPK
I don't believe the dev server is supported with 2.7 at all. That was the
position during the trusted testers phase
and all of the recent posts here by google indicate that is still the case.
Rgds
Tim
--
You received this message because you are subscribed to the Google Groups
"Google App En
HI
Umm it was requested, but also for support for urllib2 took a while.
Also urlfetch allows for async fetching which you can't do directly with
urllib, so unless people where using existing
libs (ie a soap client) you where always better off using urlfetch.
Thanks
Tim
--
You received this m
Hi
Currently you set it globally with urlfetch.set_default_fetch_deadline(n)
You can also query the current value with get_default_fetch_deadline
Rgds
Tim
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To view this discussion on the
On Wednesday, October 5, 2011 9:08:52 PM UTC+8, PatrickCD wrote:
>
> Tim,
>
> > I actually think a Expando model might be more appropriate combined with
> > adapters.
>
> Hi. As far as I understand it, Expando models allow you to store
> additional data, but I'm looking to change behaviour, n
Agreed, they could upload a new version of the app undermining anything you
might have done.
--
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/-/qgFwE
Hi
I would take Nick's advice. I have seriously hacked up remote api shell for
my own purposes.
Even going as far as using it inside Zope to seamlessly access both zope
object store and
the google datastore are the same time.
T
--
You received this message because you are subscribed to the
I agree
I actually think a Expando model might be more appropriate combined with
adapters.
That way you can store additional information if required due to different
requirements,
but keep the specific state based methods in the adapters, if you don't want
to start
aggregating methods from di
I agree, with you jeff, I don't believe google has ever made that claim.
They have released the SDK source (and NDB) and the api, but thats as far as
it has gone for appengine.
Hence we have things like typhoonae which are api compatible with appengine.
As far as I am concerned the scheduler is
Hi
> My app traffic is very small now. This is the pain.
> I just found the scheduler doesn't restart the idle instance
> automatically, or doesn't restart the idle instance when the instance
> is idle.
>
Instead, it restarts the idle instance when a visit comes. This is so
> bad. It w
Do you really think the shotgun approach of swamping the group with lots of
little
similiar postings with either just a comment/statement and sometimes a
question
will get you the information you want ?
--
You received this message because you are subscribed to the Google Groups
"Google Ap
You haven't deployed the gdata library.
It needs to be installed and deployed within your project as it is not part
of appengine.
Rgds
Tim
--
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://
Think ?
Actually run tests and measure. Too much thinking and not enough actual
testing
won't solve your problems.
I can imperically say warmup requests help on all of my applications.
T
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
Maybe you shouldn't use struts, this is a serious suggestion, and should be
considered seriously.
If you have a lightly used small app using a heavy framework then you have
probably made a bad design choice.
If you have long startup times then warmup requests do help enormously, and
in fact are
Unfortunately in all your posts, you provide little background.
So we have contextless complaints, its difficult to work out what you
have done, and I won't assume you have done things like optimisation.
Also are you handling warmup requests ?
T
--
You received this message because you are sub
I have a site, with max idle set to 1, and Min Pending Latency set to 500ms,
and I find performance is great.
It takes 3-4 secs normally to start the full app up, but through smart
caching (I can start a cold instance and serve many pages directly from
memcache without starting the full stack i
p the
entity with an adapter with the methods you want etc... based on the status.
I feel this would be a better approach. I use adapters (zope.component) on
appengine in pretty much all of my projects. It would
give you the same behaviour r without having to hack around with the
PolyModel metacl
Hi
I run a complete separate instances for dev/test/uat with their own data.
Then final testing with live data in production instances before making the
new version the default.
Rgds
Tim
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group
The slider for max idle instances now has two sliders.
I am assuming the top one is min idle instances and the bottom one is max
idle instances (which was there before)
I think the pair of sliders could be better documented in the application
settings page.
Rgds
Tim
--
You received this messa
I would strongly agree with Robert, and not agree with the OP.
I find the appengine docs are fine, everything can do with improvement
though ;-)
Just my 2c worth
T
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To view this discussion o
Probably the easiest thing to do , is upload your data to the local dev
server,
Then you can run your app, or play with the data directly with the remote
api
Rgds
Tim
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To view this discussi
passed through to
appengine apps.
Rgds
Tim
On Sep 19, 8:43 pm, Alexander Konovalenko wrote:
> On Mon, Sep 19, 2011, Tim Hoffman wrote:
> > I think google are saying they are reserved as guide so you don't clash with
> > some service of theirs, however I don't actuall
Hi Alexander
I think google are saying they are reserved as guide so you don't clash with
some service of theirs, however I don't actually see
any way of preventing you from using any of these paths in urls if you
handling them yourself.
In my case I don't use webapp and don't only map a few o
+1 too!
--
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/-/8K-4RZF_TrwJ.
To post to this group, send email to google-appengine@googlegroups.com.
To un
Hi
You could submit the request via ajax back to your appengine app
and it can then do an async requuest on all the urls, .
In your case you have some of the info already and have to fetch some of
it.
So it might be two ajax calls, one to get the list of books, the result is
book prices for st
Hi
Your handler needs to trap the exception..
The an error won't be delivered back to the mail delivery service.
It is retrying because it can't tell why it got the error from your handler.
Rgds
Tim
--
You received this message because you are subscribed to the Google Groups
"Google App Engi
HI Milosh
Unfortunately there is no guarunteed way to be sure someone has read the
mail sent.
In fact it would in some cases be considered an invasion of privacy ;-)
If's not fundamentally part of email. and read receipt implementations are
very much
mail server dependent.
The most you can sa
Hi
That will only tell you they read it, if they have images enabled by
default. I don't for instance in my email.
If they don't read it with images enabled you can't tell if it was not
delivered or just unread.
I wouldn't expect exceptions to be raised from bogus email addresses unless
they
Hey
Your right. Not good .
Here is the link to the group
https://groups.google.com/forum/?hl=en#!forum/google-appengine-python
--
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.goo
Hi
The group is google-appengine-python
Rgds
Tim
--
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/-/bmtjKeHqQ9YJ.
To post to this group, send emai
Hi
Sorry no specific links.
Have a read of appcfg.py in google/appengine/tools/appcfg.py , specifically
the class
AppVersionUpload.
It uses remote_api rpc calls under the hood
All of the appcfg functions that talk to a remote server do.
If you wanted to build your own wrapper I would be usin
HI
As robert says it is just python. Also its using the remote_api.
So all definately do-able. I run remote_api stuff inside zope to
transparently access the datastore from within zope.
Rgds
Tim
--
You received this message because you are subscribed to the Google Groups
"Google App Engin
Hi
You really need to provide more information if you want some help.
Have a look in the dashboards log for the request and see what error details
are logged.
Rgds
T
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To view this discussio
Hi
Did you know the dev SDK still doesn't support python 2.7.
(Will soon but not right now)
Have you tried running the same code on python 2.5 ?
Though its probably not the problem.
Have you actually looked at the value of val from your line
dom_val = minidom.parseString(val)
To see if it is
Hi Jon
Thanks for the update.
Faster billing will help a lot.
I don't envy the person who gets to write up all these billing behaviours in
a
clean concise fashion. A job for an good technical writer.
See ya
T
--
You received this message because you are subscribed to the Google Groups
"Go
Ok, where he said that was in the context of someone pinging every minute
rather than as
a general statement.
It would be good if the instances view showed billable instance hours
against each instance.
I might have to wait more than 3 days to see the billing, I am currently 4
days behind on
Hi Barry
Thanks for pointing out that thread.
I am still not 100% convinced from what Jon said is the whole picture
He did say that they may not kill of idle instances and the graph shows
that.
But he also says that max-idle-instances=1 equates to 24 billable hours.
So I am inclined to conc
asonable that any idle instance should be shut down after 15
minutes.
Regards
Tim Hoffman
--
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/-/
Yes
Thats what I have been trying to convey from the beginning.
warmup requests have their place, its obviously better if you apps starts up
quick enough that you don't need them
but I think they are probably few and far between ;-)
T
--
You received this message because you are subscribed to
HI Barry
Totally agree. Those are the costs that the spreadsheet is ignoring amongst
other things.
On the face of it, for small apps the actual base hosting cost of appengine
vs the
other solutions are all in the same ball park.
Statements like appengine is no good for small apps is probabl
I would think your mad to use a framework with a typical startup time for an
app instance
of 10 secs if your only going to serve 100 requests a day.
As I said the value of /_ah/warmup depends on the situation.
Tim
--
You received this message because you are subscribed to the Google Groups
That would entirely depend on how long its take your instances to start up.
If you have something really heavy like >10sec then I think /_ah/warmup is
essential
Just my 2c
T
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To view this d
HI Adam
Thats a comparison I would like to see. It still ignores large scaling
etc...
But it would provide a minimal application change option for running else
where.
Which is a much better real comparison.
Rgds
T
--
You received this message because you are subscribed to the Google Groups
Hi Strom
Yep, though accessing it counts in rpc quotas.
I tried to keep the stack really small ;-)
Honestly you can't compare any VPS with appengine feature wise unless you
build some really big
stack yourself and you still don't get the seemless scaling. But for really
small apps the scaling i
Let me know if you would like to help collate this information, and I can
give you write permissions on this
spreadsheet.
Hopefully this is a useful excercise.
Also I need to look at what the minimum spec for a typhoonae or appscale
deployment would be, if someone wanted to move
off appengine b
Hi Joshua
Thanks for your input I feel I can get my small instances down to the
$9-$15 per month territory too.
But I wanted to not have to argue that point too much here ;-)
Anyone running a small instance that is well optimised should have a good
handle on what it will cost them.
So lets r
+1 from me on that score Brandon.
T
--
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/-/WUlDa8bUNGgJ.
To post to this group, send email to google-ap
Hi Raymond
This particular discussion was focused on small apps. Not big apps. Note
the subject "GAE pricing is not suited for smaller apps"
I have been and am involved with large apps. Thats a whole different
discussion. And I have repeatedly stated I am not talking about pricing at
that sc
Hi
More than a few people have said in the groups lately that appengine is
unsuitable for entry level apps due to the new pricing schedule.
I am not so sure, but there hasn't been any real information about the
alternatives, so I thought I would start to collate some numbers
So to that end I ha
Didn't you read the going out of preview notice from google.
They garuntee to provide the service for 3 years. This gives organisations
piece of mind.
And you are drawing the conclusion from that report which is very vague,
that google plan to kill gae. What have you been smoking ?
Its compari
Why do we keep seeing these comments "Yikes, maybe Google now has too few
large customers and that they now on purpose have chosen to kill GAE"
Where is the evidence that allows you to draw that conclusion?
or are you actually trying to start or perpetuate an internet meme and hope
it comes tru
Hi Gubbi
I feel I can halve them because feel a lot more comfortable if there are no
instances hanging around and have them start on demand
on HR. M/S startup has always proven to be problematic.
I have a test instance that I check every so often on HR and it has yet to
fail to startup once or
Raymond
Unfortunately I don't see a lot of postings detailing what the real
alternatives to appengine are, what
certain size traffic sites and data sets would actually cost to run on these
alternatives.
May be you could provide a bit of an over view of your apps design, what its
costing you,
Hi Raymond,
Lets look at some real numbers rather than broad statements.
In fact in the second year an AWS micro instance is charged at 0.02 per hour
So
>>> 356 * (24 * 0.02)
170.88
a reserved instance is
>>> 356 * (24 * 0.007) + 56
115.808
vs appengine with a single instance around $3
I can host on AWS without costing a buck for the first year with a micro
instance.
Second year it will cost.
Also I need to keep ubuntu patched, my ebs volume clean (clean out logs).
etc
AWS is not immune to outages (neither is appengine)
And heroku, well lets say I am a python person thro
i never said it should contribute more to the bottom line, and there is no
evidence to suggest it does.
All of this is just supposition on your part. Do you any evidence to back
up your claims.
Enough said by me on this now.
See ya
T
--
You received this message because you are subscribed t
So if it is so easy to scale up and out with these other solutions, why are
you still here?
Why did you even start with appengine?
What value you do you see in appengine vs these other IAAS platforms?
I really would like to know. You seem to only ever have negative comments
about appengine a
Hi
I don't want to migrate because I don't want to deal with os and complete
application stack.
I don't want to have to wake and find something has fallen over.
Also I know I can reduce my cost below $30 a month, probably to half that,
when I move to HRD.
I have had that site on appengine for
I totally agree with you francois I just found the whining without
constructive dialog starting to be a more than a little painful.
I don't have a problem helping people get more out of appengine, but
baseless unsupported sniping isn't helping anyone, and the signal to
noise ratio is drowning
I never said trolling, just said it was nonsense, with no evidence to
support the claim.
In addition the basic out of the box google infrastructure doesn't do
appengine. Appengine is a whole new
layer that has been developed,therwise you wouldn't see things like M/S or
HR datastore's
And you
Don't agree, I don't see the evidence to support you claims.
I think it has probably been more successfull than they can afford and need
to start re-couping the
massive cost of running such an service.
But thats my opinion, only google can really tell us the truth of the
matter.
I will leave i
3 years is not a slow kill.
Thats like saying Canonical is deprecating ubuntu because Long Term Support
(*LTS*) version you get 3 years committed support.
What the 3 years statment provides is organisations a committment they can
rely on, which was never there
whilst appengine is in preview.
This is all just nonsense, they are changing their business model.
They wouldn't be taking appengine out of preview and then legally committing
to the platform for 3 years
if they where going to can.
--
You received this message because you are subscribed to the Google Groups
"Google App Engi
Interestingly prior to the new scheduler changes
you could always get instances to disappear (the good ol' days ;-)
Maybe moving forward we don't always get better ;-)
T
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To view this discuss
Hi
I have an unbilled HR app in development, that isn't being used much (say
one request about every few hours),
and it seems to sit for hours with a single instance idle for very long
times,
Total number of instancesAverage QPS*Average Latency*Average Memory1 total
0.000Unknown ms43.0 MBytes*
Hi
I have been thinking further about what I think the bare minimum the graphs
should be showing.
It really needs instance hours cost on 15 min increments (or less) or if
that is not possible at least instance hours,
number of instances isn't the metric we are going to be charged on.
For bur
Oh this is all nonsense.
You are all smart enough to understand how it works.
Google has dropped the ball, not providing appropriate real time tools to
monitor instance hours and what it will cost you.
You can also argue about the level of pricing, but to suggest it is any more
complicated
+1 Geoffrey
This is way to much vitriol rather than constructive dialogue in this group
at the moment.
T
--
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-app
With this code you I think you won't be reclaiming any memory.
You could try a gc.collect, in addition move the deletions in your code so
that the deletes occurr in a function.
I would have the outer loop use a cursor, pass the query to a function that
does the fetch of the keys
and does the de
Hi
Actually I believe I can get my price down below $30 a month, keeping a
single instance running will cost around $30.
I can get down to single instances no problems. Sure the scheduler isn't
helping.
Total number of instancesAverage QPS*Average Latency*Average Memory1 total
0.01761.0 ms49.9
Please note, I am not an apologist for the new billing regime, I think there
have been some major misteps with it,
especially with regards to clarity of information , lack of tools to
really gauge the impact of your own chnages.
Billing stats 3 days old is pretty useless if you trying to reduce
Hi Raymond
Here you go, posted about it earlier today.
This is at the low end of the equations and numbers talkd about around here.
But appengine still works for me.
https://groups.google.com/d/msg/google-appengine/0tKkjM5xlc4/1bmSJx5huvwJ
Rgds
T
--
You received this message because you a
Absolutely not.
I don't regret a single thing I have said about appengine. I always explain
the risks too. I have been here a long time, and been pretty vocal about
shortcomings.
T
--
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To vie
But definately the scheduler is currently screwy and needs some serious
work.
T
--
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/-/dMQ0mKkqjysJ.
To
Yep
I am sure they will have some teething issues ongoing with the scheduler.
I have left things on automatic, and and dropped always on, and almost 90%
of the time I have a single instance.
Rgds
T
--
You received this message because you are subscribed to the Google Groups
"Google App Eng
HI
I run a small business web site on appengine (since early 2009)
www.fishandlily.com.au its a dynamic website with a simple cms (also got
some big apps on other instances) I was running "always on" which was only
$9 month, which protected me from the odd slow start up time on M/S and I
did
HI Greg
Yep the graph might not be that useful, however billing stats are running at
about 3 days behind, so between late billing stats and a less
than useful graph in terms of estimating your actual costs the round trip
time on making real changes and see the true impact on real costs
is prett
Hi
All of my billing history is running at least 3 days behind actual time.
Its the 3rd of Sept, and the last Usage report I have if for the 29th of
Aug, which was generated on the 1st of Sept.
I am GMT + 8 , so things are expected to be a little out , but not that far.
This makes it very diffi
Hi
I found if you have "always o"n set in your billing changes to max idle
instances and min pending latency
can't be changed.
Rgds
T
--
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://gro
HI Barry
I realize they area under the graph is number of instances, however in my
case there is in fact a lot of instance churn (especially when I drop off
always on)
Like I said later maybe graphs aren't the best way to provide this
information.
I do want to see the rate of instance creatio
Hi
Actually I would disagree on the M/S front, all of my apps are on M/S and
even despite the recent issues
M/S is way more stable than it was in 2008 and 2009, I mean WAY MORE STABLE
;-)
Thats not to say more growing pains won't be felt. My guess is with a
supposed imminent exodus of apps, w
Just look at the stack in that example you posted. It is a non trivial
excercise and this guys has gone in deep.
How many people are willing too or can..
And it doesn't talk about how robust it is. Its not a lot of fun performing
rolling upgrades of cluster of
rdbms backends whilst keeping ev
Hi Greg
You didn't really read my post. I explicitly said the graph of instances
doesn't really help.
How do you guage you instance hours accumulated for the hour from the
instances graph that is currently there.
Maybe a graph isn't the right tool, Can you tell me how many instance hours
and
Hi Raymond
You will need more than a single sysadmin to make a system scalable enough
on AWS
to deal with a royal-wedding-bells-in-cloud and even on AWS it will cost you
a lot.
Not defending in anyway price changes, but you are seriously over
simplifying what is required to build
a large scal
1 - 100 of 757 matches
Mail list logo
