Do these problems appear only with threadsafe on? Or do they only occur when
threadsafe is set to off?
Ikai Lan
Developer Programs Engineer, Google App Engine
Blog: http://googleappengine.blogspot.com
Twitter: http://twitter.com/app_engine
Reddit: http://www.reddit.com/r/appengine
On Thu, Apr
I disabled threadsafe and AlwaysOn and it got worse. Many 500 errors on
instance startups, all with this message:
The process handling this request unexpectedly died. This is likely to
cause a new process to be used for the next request to your
application. (Error code 203)
When I re-enabled
Wow, thanks for the feedback. Does anyone here have a negative experience
with thread-safe?
I'm asking because we had received some feedback that this resulted in
poorer performance. We don't believe this should be the case at all and want
to track down if and why this happens.
One huge benefit
Hi,
It all appears very good so far from my experience - I haven't personally
noticed any drop-off in performance, only a massive decrease in the number
of instances spun-up and therefore a large drop-off of warm-up requests. A
big win!
Cheers,
Simon
--
You received this message because
Hey Ikai,
Thanks for the explanation about the pending queue, I think that may
be the most information I've ever seen about how that process works.
It would be great to have that information included somewhere in the
docs. At the minimum it can help understand certain behaviors we see
Hi Ikai!
I didn't notice any performance issue here.
However I'm experiencing a lot of errors lately during instance startup.
I've AlwaysOn and the Instances panel clearly shows that 2 instances are
always ok, with a large Age. But one of those instances is constantly being
shut down and
Are you using Master/Slave or HR?
2011/4/7 Sérgio Lopes slo...@gmail.com:
Hi Ikai!
I didn't notice any performance issue here.
However I'm experiencing a lot of errors lately during instance startup.
I've AlwaysOn and the Instances panel clearly shows that 2 instances are
always ok, with a
Master/Slave. But the exceptions don't seem to be related do datastore. Most
errors are with /_ah/warmup requests, and my startup process doesn't touch
the datastore.
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To post to this group,
I've been using GAE 1.4.3 with thread-safe on and AlwaysOn for a few days.
And my conclusion is that it almost solves all Cold Start problems we were
used to. Now I have fewer instances warmups and most of them are /_ah/warmup
request with no user suffering. In my tests, I had instances lasting
I tried putting threadsafe: true in app.yaml and it worked. I put it as so:
application:
version: 1
runtime: java
threadsafe: true
snip
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To post to this group, send email to
When you say it worked, how did you get confirmation? Is there a message,
or did it simply not throw an error?
--
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
Ah ignore my message, I've just tested it - in SDK 1.4.2 it throws an error
and in SDK 1.4.3 it doesn't. Cheers for the info, I should really have
guessed at that...
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To post to this group,
What dies thread safe really mean ? (your application code needs to
use proper thread synchronization before you enable threadsafe).
Can someone give some hints?
On Mar 31, 10:53 am, Simon Knott knott.si...@gmail.com wrote:
Ah ignore my message, I've just tested it - in SDK 1.4.2 it throws an
On Thu, Mar 31, 2011 at 1:38 AM, JH jamesthop...@gmail.com wrote:
Been looking forward to files api for a while!
I have been playing with it and it seems there is an issue, I can't
figure out if it's me or the files api...
I kick off tasks which use the api to write about 300 files. It
Yes, I did. about 10-25% of the time it would fail when using
blob_key to set a blobreference. I added a time.sleep(1) and problem
solved, so appears to be some type of consistency issue with the
blobinfo ??
On Mar 31, 6:58 am, Stephen sdeasey+gro...@gmail.com wrote:
On Thu, Mar 31, 2011 at
Hello,
* - You can now configure the specific application version to which a task
queue*
* or cron job will send requests.*
How use this functionality?
Thank you
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To post to this group,
http://code.google.com/appengine/docs/python/config/cron.html#Cron_and_App_Versions
On Thu, Mar 31, 2011 at 11:05, Sahid Orentino Ferdjaoui
sahid.ferdja...@gmail.com wrote:
Hello,
- You can now configure the specific application version to which a task
queue
or cron job will send
Ice13ill,
There's a short and a long long answer to your question.
The short one: If you don't already know what proper thread
synchronization is, leave the threadsafe option switched off.
The long one: Switching it on allows multiple HTTP requests to be
handled concurrently by the same
Andrei, that's probably the right implementation, considering that an
application can run in multiple JVMs.
Where this sort of thing becomes very tricky is when people start using the
JVM's local memory as an even faster cache than Memcache and don't
synchronize access to data structures, which
Well, i my servlet methods (i use GWT RPC) do not use Thread specific
operations and they don't access any shared objects other than
Memcache objects, Session, or Datastore entities.
On 31 March 2011 18:20, Remigius remigius.stal...@gmail.com wrote:
Ice13ill,
There's a short and a long long
Hi,
Just want to verify... the Prospective Search API is NOT the same
thing as the full text search API thats mentioned in the roadmap, is
that right?
I certainly hope so, as I very much need Retrospective and not
Prospective search.
--
You received this message because you are subscribed to
What do Prospective and Retrospective searches refer to ?
On 31 March 2011 20:31, tempy fay...@gmail.com wrote:
Hi,
Just want to verify... the Prospective Search API is NOT the same
thing as the full text search API thats mentioned in the roadmap, is
that right?
I certainly hope so, as I
Is the JavaDoc for the Files API online somewhere? I noticed the Blobstore
Overview was updated to include some instructions (
http://code.google.com/appengine/docs/java/blobstore/overview.html#Writing_Files_to_the_Blobstore),
but I couldn't find the JavaDoc either in the SDK download or the
Prospective search is not the same as full-text search.
You can read about it:
http://code.google.com/appengine/docs/python/prospectivesearch/
On Thu, Mar 31, 2011 at 13:31, tempy fay...@gmail.com wrote:
Hi,
Just want to verify... the Prospective Search API is NOT the same
thing as
Been looking forward to files api for a while!
I have been playing with it and it seems there is an issue, I can't
figure out if it's me or the files api...
I kick off tasks which use the api to write about 300 files. It seems
about 10% of them fail, or when I run:
blob_key =
FYI: To fix this I had to add a time.sleep(1) before setting my
blobreference property = to blob_key =
files.blobstore.get_blob_key(file_name) ...
maybe it takes a few hundred milliseconds for this to be available? or
some type of consistency issue when setting the reference property?
On Mar 30,
Is there anyway to set the *threadsafe* setting using the YAML configuration
for a Java app?
I could see that the appengine-web.xml docs had been updated, but can't see
a similar update to the YAML config page.
--
You received this message because you are subscribed to the Google Groups
27 matches
Mail list logo
