I would like someone to clarify something for me:
There are two kinds of timeouts that exist during SOAP requests/responses:
1- Connection Timeout
2- Read Timeout
This applies at least to Axis1/Axis2, which I’m currently using.
The connection timeout happens when the client couldn’t connect to the web service in question within the set Connection Timeout value, and which would eventually result in throwing the following exception :
Could not connect to host within a timeout of “value”.
As for the Read Timeout, I’m really not sure about it, and I don’t know which assumption is true. Let’s take a scenario for example, in which a client is sending data to a web service, which will in turn process the data, checks for their sanity, inserts them into the database when they are, and then the web service will send some data back to the client. Bottom line, we have a significant amount of processing time on the server, and significant data that’s being sent back and forth between the client and the web service.
What I’m unable to understand is when is a read timeout exception thrown by the client?
1- Could it happen when the client is still in the process of marshaling the objects that are being sent to the web service?
2- Could it happen during the process when the web service has already started writing its response to the open socket?
I could really appreciate clear answers on this. Thanks a lot in advance.
It’s much clearer now thanks to the efforts I did to research this. A “Read Timeout” is basically when the client hasn’t gotten anything byte of date still. So let’s take a scenario where a server needs to reply back to a client with 4 MBs of data. Read Timeout will be reset with every byte of data the client is receiving from the server.