[Apiman-user] service not being published to gateway after restart?

[Jakub Čecháček]

Hello,

rm ${APIMAN_HOME}/standalone/deployments/apiman-es.war

Removing the deployment is all you need to do.

Jakub.

On Wed, Oct 14, 2015 at 11:57 AM, Tim Dudgeon <tdudgeon.ml at gmail.com> wrote:

> OK, I tried that and now the gateway is working after a restart.
> I have no idea why is didn't work when running with the built in ES.
>
> Now one more question - how to disable the ES within apiman (I mean stop
> it running, not reconfigure the connection)?
>
> Tim
>
> On 13/10/2015 14:50, Eric Wittmann wrote:
> > Since you have to do it anyway, I'd suggest going with an external
> > elasticsearch instance.  It's easy to hook apiman up to it.  It's just
> > a couple properties in apiman.properties.
> >
> > -eric
> >
> > On 10/13/2015 9:48 AM, Tim Dudgeon wrote:
> >> If you restart a container it will contain the same state as when it was
> >> stopped.
> >> If you were to create a new container from the image it would of course
> >> not contain the state of other containers created from that image
> >> (unless you were re-using volumes).
> >>
> >> And with the jboss/apiman-wildfly image its working OK. I stop the
> >> container and restart it and the gateway is configured.
> >> But in my image, adapted from that image the gateway looses it
> >> configuration, so its clearly something I've done. I just don't know
> >> what.
> >>
> >> What I've discovered so far is that when restarted ES contains
> >> information in the apiman_manger index but has lost all info in the
> >> apiman_gateway index.
> >>
> >> Tim
> >>
> >>
> >> On 13/10/2015 13:44, Marc Savy wrote:
> >>> Not a container expert either, but:
> >>>
> >>> I think it's standard practice for containers to be non-persistent
> >>> (i.e. no persistent state), so you'd need an external ES instance. or
> >>> set the ES data-store to be on a data volume (which is persistent).
> >>>
> >>> Regards,
> >>> Marc
> >>>
> >>> On 13/10/2015 12:57, Eric Wittmann wrote:
> >>>> Hm.  Without digging into it, I'm not sure what to suggest.  I'm also
> >>>> not a docker expert by any stretch, so that complicates things. :)
> >>>>
> >>>> I can tell you that the default elasticsearch instance that comes with
> >>>> the apiman quickstart stores data in the standalone/data/es directory.
> >>>> You could perhaps poke around that area.  See if data is getting
> >>>> written
> >>>> there, and whether it's still there when you restart the container?
> >>>>
> >>>> -Eric
> >>>>
> >>>>
> >>>> On 10/13/2015 7:37 AM, Tim Dudgeon wrote:
> >>>> > Right now I'm using the ES that's bundled with Apiman (though do
> >>>> intend
> >>>> > to switch to a separate container).
> >>>> > The baffling thing is that with the Docker image on Dockerhub
> >>>> > (jboss/apiman-wildfly) it restarts OK, but with my partly Dockerised
> >>>> > version (with Keycloak and Postgres currently as separate
> >>>> containers) it
> >>>> > fails. The APIMan and ES parts are still pretty well the same as the
> >>>> > jboss/apiman-wildfly image.
> >>>> >
> >>>> > So I seem to have messed something up, but don't have much of an
> >>>> idea as
> >>>> > to what.
> >>>> >
> >>>> > Tim
> >>>> >
> >>>> > On 13/10/2015 12:23, Eric Wittmann wrote:
> >>>> >> When a service is published to the gateway, that configuration is
> >>>> >> saved in the gateway's persistent registry.  By default (in the
> >>>> latest
> >>>> >> version of apiman) the gateway uses elasticsearch to store its
> >>>> >> registry information.  So ideally you should run elasticsearch as a
> >>>> >> separate docker container, making sure that you have an appropriate
> >>>> >> persistent volume attached so that any data added to ES survives a
> >>>> >> docker restart.  Then you can configure apiman to use that
> >>>> external ES.
> >>>> >>
> >>>> >> There is documentation in the Installation Guide that explains
> >>>> how to
> >>>> >> use an external elasticsearch cluster.
> >>>> >>
> >>>> >> -Eric
> >>>> >>
> >>>> >> On 10/11/2015 7:30 AM, Tim Dudgeon wrote:
> >>>> >>> I'm hitting a problem where I'm running apiman in a docker
> >>>> container.
> >>>> >>> When I first set up a public service it runs fine.
> >>>> >>> But when I restart the container the service definition is still
> >>>> there,
> >>>> >>> but the gateway does not respond to requests for the service. If I
> >>>> >>> create a new version of the service and publish it then it works
> >>>> fine.
> >>>> >>>
> >>>> >>> Seems like the service definition is not being published to the
> >>>> gateway
> >>>> >>> after a restart?
> >>>> >>> I've tried it with the basic Docker container similar to
> >>>> >>> jboss/apiman-wildfly on Dockerhub and it seems OK with that, so
> >>>> it looks
> >>>> >>> like its something I've done to cause this, but I don't see any
> >>>> errors
> >>>> >>> and don't know where to look to try to fix it.
> >>>> >>> Any ideas what's wrong?
> >>>> >>> This is with version 1.1.8.Final.
> >>>> >>>
> >>>> >>> Tim
> >>>> >>>
> >>>> >>> _______________________________________________
> >>>> >>> Apiman-user mailing list
> >>>> >>> Apiman-user at lists.jboss.org
> >>>> >>> https://lists.jboss.org/mailman/listinfo/apiman-user
> >>>> >>>
> >>>> >
> >>>> > _______________________________________________
> >>>> > Apiman-user mailing list
> >>>> > Apiman-user at lists.jboss.org
> >>>> > https://lists.jboss.org/mailman/listinfo/apiman-user
> >>>> >
> >>>> _______________________________________________
> >>>> Apiman-user mailing list
> >>>> Apiman-user at lists.jboss.org
> >>>> https://lists.jboss.org/mailman/listinfo/apiman-user
> >>>>
> >>>
> >>
>
> _______________________________________________
> Apiman-user mailing list
> Apiman-user at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/apiman-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/apiman-user/attachments/20151014/c409a7b0/attachment.html