ant-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Conor MacNeill" <co...@cortexebusiness.com.au>
Subject Re: Performance with <wlrun>
Date Tue, 24 Apr 2001 11:29:18 GMT
David,

I never use wlrun in a production environment.

There could be two factors slowing down wlrun. One would the presence of
the Ant JVM process and the memory load that may present to the system and
the other, more likely, would the Stream Pumpers used to pump the weblogic
output back into Ant's event output. If you are producing a lot of output,
it will be going through the pumpers' small buffers with a built in 5 msec
sleep delay (at least) every 128 bytes. That translates to about 25 k/sec
which isn't a lot of output. Above this rate, something will eventually
block. Are you producing a lot of output during deployment. I wouldn't have
expected that to be the case.

Not sure of the need for the sleep since the read is blocking anyway.

Conor

----- Original Message -----
From: "Ventimiglia, David" <David.Ventimiglia@morganstanley.com>
To: "Ant-User (E-mail)" <ant-user@jakarta.apache.org>
Sent: Tuesday, April 24, 2001 9:10 AM
Subject: Performance with <wlrun>


> Hi,
>
> Weblogic is much slower to deploy EJBs when I use <wlrun> than when I use
a
> simple script to start Weblogic.  Are there any known performance issues
> with the <wlrun> task?  Has anyone done any kind of analysis on this?  Is
> anyone using the <wlrun> task in a production environment, or is it only
> being used during development?  I ask because in general, <wlrun> is much
> nicer than hairy startup scripts and I see no obvious reason to not use
it
> in a production environment, EXCEPT that it appears slow (but then maybe
I'm
> not using it correctly).  Thanks!
>
> Cheers,
> David
>


Mime
View raw message