[Odlparent-dev] [netconf-dev] [release] netconf SFT hanging for over 3 hrs


Robert Varga
 

On 20/07/17 16:35, Ryan Goulding wrote:
It fixes it in a bandaid sort of way... the correct solution is to move
to Jersey 2 so that jersey instantiates these objects through java code
and not web.xml at runtime. Then we can do correct bean instantiation
and wiring. There is a separate stopgap I was investigating through
using a web.xml Listener but it appears that doesn't work. This at
least gets us over the initial bug, but the long term solution is to use
Jersey 2.
Right. We need to find someone who will drive the upgrade -- we have
been talking about it for 2+ years now
(https://bugs.opendaylight.org/show_bug.cgi?id=2515) but cannot seem to
make any forward progress :(

Regards,
Robert


Colin Dixon <colin@...>
 

What's blocking it? Or do we not know?


On Thu, Jul 20, 2017 at 1:23 PM Robert Varga <nite@...> wrote:
On 20/07/17 16:35, Ryan Goulding wrote:
> It fixes it in a bandaid sort of way... the correct solution is to move
> to Jersey 2 so that jersey instantiates these objects through java code
> and not web.xml at runtime.  Then we can do correct bean instantiation
> and wiring.  There is a separate stopgap I was investigating through
> using a web.xml Listener but it appears that doesn't work.  This at
> least gets us over the initial bug, but the long term solution is to use
> Jersey 2.

Right. We need to find someone who will drive the upgrade -- we have
been talking about it for 2+ years now
(https://bugs.opendaylight.org/show_bug.cgi?id=2515) but cannot seem to
make any forward progress :(

Regards,
Robert


Ryan Goulding <ryandgoulding@...>
 

Lack of volunteers.  Its a non trivial upgrade.

Regards,

Ryan Goulding

On Wed, Jul 26, 2017 at 8:56 AM, Colin Dixon <colin@...> wrote:
What's blocking it? Or do we not know?

On Thu, Jul 20, 2017 at 1:23 PM Robert Varga <nite@...> wrote:
On 20/07/17 16:35, Ryan Goulding wrote:
> It fixes it in a bandaid sort of way... the correct solution is to move
> to Jersey 2 so that jersey instantiates these objects through java code
> and not web.xml at runtime.  Then we can do correct bean instantiation
> and wiring.  There is a separate stopgap I was investigating through
> using a web.xml Listener but it appears that doesn't work.  This at
> least gets us over the initial bug, but the long term solution is to use
> Jersey 2.

Right. We need to find someone who will drive the upgrade -- we have
been talking about it for 2+ years now
(https://bugs.opendaylight.org/show_bug.cgi?id=2515) but cannot seem to
make any forward progress :(

Regards,
Robert