On 03/07/2014 16:03, Christopher Schultz wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
Paul,
On 7/2/14, 4:28 PM, Paul Taylor wrote:
On 02/07/2014 16:34, Christopher Schultz wrote:
The solution is that the web application, packaged in a WAR
file, needs to unpack the Lucene indexes onto the disk when it
starts up. You can do this with a ServletContextListener.
So I do within init() method of my servlet, but EB doesnt wait for
the init() method to finish before declaring the application ready,
do you think it would wait for code using a ServletContextListener
or fail in the same way it does for init() ?
Which init() are you talking about?
GenericServlet.init(), is this method not designed for what I was doing
If EB marks the cluster node as
available before the webapp has completed deploying, that sounds like
a big problem. If you are using servlet.init() without any
load-on-startup for that servlet, then I would expect the bad behavior
you describe.
I dont understand what you are getting at here when you say
'load-on-startup'
I think you'll have better luck with
ServletContextListener, because the servlet spec guarantees those to
be complete before the webapp can receive requests.
Okay, well the good news is that I now have it working with AWS
ebextensions and it wasn't so difficult in the end.
You want to manage the streams yourself: hook a gzip reader up to a
tar reader and then read files from the tar reader:
GzipInputStream gzin = new GzipInputStream(tgzFile);
TarInputStream tin = new TarInputStream(gzin);
InputStream fin;
while(fin = tin.getNextFile())
copy(fin, new File(localFilename));
That's oversimplified, but essentially what you want to do. There is
no need to completely expand the gzip archive before reading files out
of it.
Ok, thanks good to know
thanks for your help
Paul
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org
