hi all, I have apache httpd in front of my jboss 7.1.1. my jee webapp seems to be working great through httpd/mod_jk/jboss7 except when

i try to download a > ~500K excel document  (xml transformed via xslt). small transformations/downloads work fine.

traffic goes through apache httpd and jboss 7.1.1 is configured to use mod_jk/ajp13

It all seems to be working great, except when i hit the size of any given reports grows, then

the time to download the data is unbearable. so as the size of the file returned

through apache/mod_jk/ajp13 grows, the time it takes to get the response in the browser becomes

unusable.

if I hit jboss directly, using port in URL, bypassing httpd/mod_jk,

I can download my excel document via a xslt transformation, any size, they all work

so opening a URL like this in browser http://myhostname.com:8080/myWebApp/TestServlet

pops the Open/Save dialog and I can see the excel document

if I hit this URL (which goes through httpd/mod_jk/ajp13) http://myhostname.com/myWebApp/TestServlet

the browser hangs only on large downloads, it never display the Open/Save Dialog, If i kill apache on the backend, then i

get the open/save dialog popping with no content in my excel doc.

if I put mod_jk in debug mode and see the transformed file being dumped to the console, on excel files returned

where size > 500Kish, the dumping of the contents goes on forever, unless i kill apache httpd.

it may return if given enough time, by is not usable.

All other files (including small excel docs) are served up no problem through httpd/mod_jk/ajp to jboss

does anyone have  jboss7/ajp13 connector experience that might give me a place to start looking?

I attached a simple TestServlet that opens xml, transforms it and sends it back to client, this

breaks with mod_jk/ajp13/jobss7.1.1