Hello;
I think Art is right here; the instance takes too long for the
response to come back so the adaptor disconnects from the instance
(causing the broken pipe error) and then tries another instance.
It is a tricky one in the scope of RPC against an application server.
In the few cases where a request would be more likely to be > ~5-8
seconds in my JSON-RPC system, I use a background thread and the
client has to poll the instance to keep checking to see if the process
has completed. This works well and largely resolves the broken pipes.
cheers.
Well, I've seen this problem when I'm testing and I'm waiting
patiently :) to get a response from an xml rpc service call.
...
Or it could be the WO HTTP adaptor timing out when it doesn't
receive a response from the WO app instance within the "receive
timeout" period (30 seconds, by default). The HTTP adaptor will
return a "No instance available" response when this occurs. The
"receive timeout" can be increased in the "Load Balancing and
Adaptor Settings" for the app in JavaMonitor.
___
Andrew Lindesay
www.lindesay.co.nz
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list ([email protected])
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com
This email sent to [EMAIL PROTECTED]
