About this I must disagree. Everything I've learned about the
appengine datastore says that unless you specifically need
transactions, you should avoid using parent entities.
When any entity is written, the optimistic concurrency journal is
maintained for the root of the entity group. If
On 6 Mar 2010, at 17:35, Nacho Coloma wrote:
-- Nacho.
P.S.: hey GAE guys: can you add a datastore.getActiveTransactions()
method? We are storing the same info using a redundant ThreadLocal!
(GAE is doing the same, just make that information public :)
Nacho, there is already a public
Nacho, there is already a public DatastoreService.getActiveTransactions()
that returns the threads open transactions. Is this not what you want?
http://code.google.com/appengine/docs/java/javadoc/com/google/appengine/api/datastore/DatastoreService.html#getActiveTransactions()
OK, now I am
Is it possible that DN L2 caching in GAE drives down CPU milliseconds
used but overall response times get slower?
Benefits of an L2 cache are subject to what you're doing (it's benefit
is for multiple PM's for a PMF accessing the same objects, and hence
providing a quick way of accessing them).
ha ha perhaps you could have designed the API... at least it shows you
are thinking along the same lines as the Google Engineers :)
On 6 Mar 2010, at 18:01, Nacho Coloma wrote:
Nacho, there is already a public
DatastoreService.getActiveTransactions()
that returns the threads open
I am doing a project where I will try to optimize google engine by using
multi-core cpus. So, now i am doing a research how the low API works in
order to find out if such an optimization is possible.
On Fri, Mar 5, 2010 at 7:20 PM, John Patterson jdpatter...@gmail.comwrote:
Basically each
Then it seems that App Engine is not the tool you want to use for this
research.
You may try paid services from other vendors, which generally would
have more resources available to you.
On Mar 6, 10:02 pm, wiilycy will...@googlemail.com wrote:
I am doing a project where I will try to optimize
Just in cast I want to make use of some existing Java APIs for messing
with DataStore, which may be the low-level API would work the best.
I just wonder, as I am still a college student, one of my lecturer
told me that we should use JPA (2.0) now as JDO is not the preferred
way for persisting
I think the gap here is that a RDBMS (PostgreSQL) and the App Engine
Datastore are totally different. In fact, I think the latter
influences your design much more than the former.
The sorts of joins you used to be able to do in your DB to efficiently
retrieve data don't work in the GAE
Since you're a student, and if you have the time, I'd recommend
learning JDO for the following reasons:
1. You already know JPA. Might as well learn something new. :)
2. JDO, imo, fits the GAE datastore much better (I think there are
quite a few discussions around here as to why).
Enjoy!
On Mar
Under 1.2.8 and 1.3.0 I created a SocketLogger that was extremely
useful in sending messages to a second machine in my development
enviroment so I could easily see what was going on in the AppServer
handlers.
Previously under Eclipse I put a jar in plugins/version-str/appengine-
Relational theory, SQL and JDBC are functional specifications and RDBMS
vendors implement these specifications. Actually speaking data model has
nothing to do with technology. A data model in software
engineeringhttp://en.wikipedia.org/wiki/Software_engineering is
an abstract model
I was looking for a way to persist my MemCache so I could do
production like testing in development mode to make sure I correctly
handle various use cases.
I tried writing a file using the technique described in
Hello,
I have one chron job that should be triggered every minute once. I
have now the problem that about every 5 minutes the following warning
occurs. I also notice that the datastore gets stuck sometimes leading
to a timeout. The later happens maybe every hour or something once.
Ther is really
SOLUTION:
I wrote it to the Datastore myself using the low-level API.
But socket I/O would be nice in the development environment.
Steve Pritchard
On Mar 6, 12:06 pm, Steve Pritchard steve...@gmail.com wrote:
I was looking for a way to persist my MemCache so I could do
production like testing
I am seeing some transaction-related behaviour, using JDO and GAE
1.3.1, that puzzles me.
I get an exception if I do something like the following. All entities
below are in separate entity groups:
PersistenceManager pm = PMF.get().getPersistenceManager();
//... fetch entity a1 of kind A,
I have the following 3-part owned relationship...
Users, the root entity, have a collection of Decks, as such (I am not
including the specific subclass of User as it doesn't seem to be
relevant):
@PersistenceCapable(identityType = IdentityType.APPLICATION)
@Inheritance(strategy =
Ok so either I'm doing something really stupid or I can't get a simple
task to execute. I have defined a task worker at /tasks/foo.
I can hit /tasks/foo no problem and the code executes fine (just a
simple servlet of course). The problem is when I add a task to the
queue like:
hi,
All emails sent from my App to free.fr domain are never received. I
received the following response from their server:
IP 209.85.221.110 is blacklisted for 85775s (too many errors
(unexisting recipients, broken connections), rejecting mail)
On other domains, all emails are received.
How I can
I'm trying to update a simple boolean value on a collection of objects
returned by a query. The objects are being returned, and I set the
value on the bean, then persist, but it's not working. What am I
missing?
The Submission class has a notifyAfter as a Long (raw date), and an
isNotified
Hi,
I've experienced the same phenomenon so that I added a new issue for
this:
http://code.google.com/p/googleappengine/issues/detail?id=2924
Thanks,
Kaz
On Feb 26, 8:05 am, Jerome jerome.mou...@gmail.com wrote:
Hello,
We have been using task queues for a few months without problem. We
21 matches
Mail list logo