Have Geek Brain

Hashtable sessions for Android

I am having a weird problem with serializing a Hashtable. I have created a Server, Client application. Where server (PC / MAC) serializes Hashtable and sends it to client (Android) over UDP. The data is being sent / read correctly, but I am getting a bunch of these messages below in the LogCat.

04-12 11:19: 43.059: DEBUG / dalvikvm (407): GetFieldID: Unable to find field Ljava / util / Hashtable; .loadFactor: F

Sometimes I saw these

04-12 11: 21: 19.150: DEBUG / dalvikvm (407): GC freed 10814 objects / 447184 bytes in 97ms

The app will run for 2-3 minutes and then crash. Interestingly, I don't see any Loadfactor errors on SDK 1.5. But I see GC Free xxxx objects that are often quiet.

After debugging, I found that the problem is with de-serialization and the error / warning comes from the following code

code:

ByteArrayInputStream bis = new ByteArrayInputStream(bytes);           
ObjectInputStream ois = new ObjectInputStream(bis);             
object = ois.readObject();

in Code:

object = ois.readObject();

on the client. My server serializes code like this.

code:

ByteArrayOutputStream bos = new ByteArrayOutputStream(); 
ObjectOutputStream oos = new ObjectOutputStream(bos);               
oos.writeObject(obj);

Any idea what's going on?

Thanks for the help!

+2


a source to share


2 answers


Don't use serialization between architectures. There is no guarantee that the serialized Dalvik VM object tree will be in a JavaSE / JavaEE compatible format. Use XML, JSON, protocol buffers, Thrift, etc. to transfer structured data between architectures.



+1


a source


Too late, but for anyone else who comes across this:

Noting what was said above (don't say you weren't warned):



I have dealt with this using protocol libraries and using maps with long objects. I understand that if dalvik or Java changes the serialization I'm full, but serialization is much easier than the alternatives, and that the error message (while benign) means a lot of unnecessary allocations and overhead. Also, I am using proguard to reduce the library to 45K (out of ~ 1M) which I need. I'd also like to suggest that you keep a copy of the trove source code in case they change their serialization, and also note that I need 3.0.0 rc1 to get it working (not sure if the serialid necessarily matches between versions).

0


a source






More articles:


All Articles