10:52 a.m.
For the web console it would be really beneficial to include the map/reduce proposal in
the next major release (Wildfly 9). Are there any objections if we move ahead prepare a PR
that provides an additional top level operation with the functionality described in
Haralds previous email?
From my point of view it should well encapsulated in a distinct
operation handler and thus I don't expect much side effects.
Regards, Heiko
1:54 p.m.
On Tue, Nov 11, 2014 at 10:52 AM, Heiko Braun <hbraun(a)redhat.com> wrote:
Are there any objections if we move ahead prepare a PR that provides
an
additional top level operation with the functionality described in Haralds
previous email?
Sure, go ahead, just not sure what do you mean by top level operation? As
code in Harald's example was using client apis and not server side ones.
But beyond that it looks fine, we would probably just want to mark that
operation as private so it won't show by default in CLI.
--
tomaz
2:07 p.m.
yes, but the code Harald demonstrated was intended to be used behind a root level dmr
operation, meaning that we intend to register an operation handler with the root resource.
similar to the "composite" operation handler.
On 11 Nov 2014, at 13:54, Tomaž Cerar <tomaz.cerar(a)gmail.com> wrote:
Sure, go ahead, just not sure what do you mean by top level
operation? As code in Harald's example was using client apis and not server side
ones.
But beyond that it looks fine, we would probably just want to mark that operation as
private so it won't show by default in CLI.
2:19 p.m.
Sure, makes sense.
On Tue, Nov 11, 2014 at 2:07 PM, Heiko Braun <hbraun(a)redhat.com> wrote:
yes, but the code Harald demonstrated was intended to be used behind
a
root level dmr operation, meaning that we intend to register an operation
handler with the root resource. similar to the "composite" operation
handler.
On 11 Nov 2014, at 13:54, Tomaž Cerar <tomaz.cerar(a)gmail.com> wrote:
Sure, go ahead, just not sure what do you mean by top level operation? As
code in Harald's example was using client apis and not server side ones.
But beyond that it looks fine, we would probably just want to mark that
operation as private so it won't show by default in CLI.
5:14 p.m.
On 11/11/14 10:52, Heiko Braun wrote:
For the web console it would be really beneficial to include the map/reduce proposal in
the next major release (Wildfly 9). Are there any objections if we move ahead prepare a PR
that provides an additional top level operation with the functionality described in
Haralds previous email?
>From my point of view it should well encapsulated in a distinct operation handler and
thus I don't expect much side effects.
Since this came up last week i've been looking a bit into whether we can
make wildcard operations working properly across processes. The main
issue seems to be some details when we determine the operation routing
in the domain.
I assume registering the operation at the root and passing the address
as a separate parameter (not called address) is mainly necessary because
of issues we are having executing composite operations targeting
wildcards in the domain?
I need to do some more investigation, but maybe i can help you with some
details getting this to work, so that the map/reduce operation would
work similar to the read-resource operation. I guess this won't change
too much of your work, it would however allow us to be more consistent
with other operations and reuse the "address" param.
Emanuel
6:09 p.m.
Ok, i'll try to catch you on irc this week to discuss the next steps
Am 11.11.2014 um 17:14 schrieb Emanuel Muckenhuber
<emuckenh(a)redhat.com>:
> On 11/11/14 10:52, Heiko Braun wrote:
>
> For the web console it would be really beneficial to include the map/reduce proposal
in the next major release (Wildfly 9). Are there any objections if we move ahead prepare a
PR that provides an additional top level operation with the functionality described in
Haralds previous email?
>
>> From my point of view it should well encapsulated in a distinct operation handler
and thus I don't expect much side effects.
Since this came up last week i've been looking a bit into whether we can make
wildcard operations working properly across processes. The main issue seems to be some
details when we determine the operation routing in the domain.
I assume registering the operation at the root and passing the address as a separate
parameter (not called address) is mainly necessary because of issues we are having
executing composite operations targeting wildcards in the domain?
I need to do some more investigation, but maybe i can help you with some details getting
this to work, so that the map/reduce operation would work similar to the read-resource
operation. I guess this won't change too much of your work, it would however allow us
to be more consistent with other operations and reuse the "address" param.
Emanuel
3462
days inactive
3462
days old
5 comments
3 participants
participants (3)
-
Emanuel Muckenhuber -
Heiko Braun -
Tomaž Cerar