Hi Bruce, 1. Why do you want to transport all the data to the front end in the first place? usually, a user can take only so much data at a time. How about using paging?
2. If you request data from the back end, you get all the results you asked for, so your approach #1 is the common scenario. If you apply paging here, there should be no problem anymore. 3. Your approach #2 is also possible. You can use channels to do what you think about. A Channel is basically an abstraction for a polling front end. The programming model allows you to realize a "push" server communication. Whether that is a good idea is up to you ;-) -> see http://code.google.com/intl/de-DE/appengine/docs/java/channel/overview.html greetz :ollix On 10 Jun., 12:18, Bruce Aloe <brucea...@gmail.com> wrote: > Hello, > > I have a system with client-server architecture. The client is > actually some java code and the server is some python code running in > GAE. When i need to query the kind (table) i defined in GAE data > store, i actually call a java function and it takes the query (e.g > select * from Person where age < 90) and send the query to the server > (some python code) through a http request. The server get and run the > query and finally generate the whole result tuples. Then the whole > result tuples will be sent back by server through a http response to > the client. Finally the client will print out the result tuples. > > As you can see, the whole system architecture is just that the client > sends a http request with query and the server returns the query > result in just a http response. However, this will not work when the > result tuples are too big (e.g 500MB data) so that a http response can > not bring the whole result tuples back to the client. > > I have two ideas for this problems: > 1 Let the client recursively sending the query and cursor information > through a http request to server and the server generates partial > result tuples for every http request and returns the partial result > tuples back to the client through a http response. As you can imagine, > there will be many http requests sending from client and many http > response sending back from the server. > > 2 Let the client sending the query only once to server and the server > recursively generate partial result tuples each time and send these > partial result tuples back to client. As you can imagine, there will > be just one http request sending from client and many http response > sending back to the server. Actually the server streams the partial > result tuples to the client. > > I prefer to implement the second idea. However, the second idea seems > impossible to be implemented because the GAE document says "App Engine > does not support sending data to the user's browser before exiting the > handler. Some web servers use this technique to "stream" data to the > user's browser over a period of time in response to a > single request. App Engine does not support this streaming technique." > The quote is fromhttp://code.google.com/appengine/docs/python/runtime.html. > > Now i can only implement the first idea. > > Do you have other good suggestions, ideas for me? > > You can of course correct me if you feel i misunderstand the GAE > document that i quote. > > Thank you very much for your valuable views. > > Bruce -- 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 this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
