Yes, to add to Alexis'
  Good question Brian, what I usually do is a ./deleteAll on everything - under 
the covers a kubectl delete
  Quick check on the Kubernetes cheat sheet docs - looks like the use of --all 
will do this
https://kubernetes.io/docs/reference/kubectl/cheatsheet/

  But restarting in this case is really delete/create - as we are stateless 
processes - were we run one app per container.  For the app server ones, you 
could interpret a replaced war as a shallow restart when tomcat or jetty picks 
up the war change and redeploys in inside a container.  But usually we just 
stop/start the container. 

  We need a minimum subset defined - there is a jira on this - to reduce the 
restart time for all of onap.
  /michael

-----Original Message-----
From: [email protected] 
[mailto:[email protected]] On Behalf Of Alexis de Talhouët
Sent: Wednesday, February 7, 2018 09:56
To: BRIAN D FREEMAN <[email protected]>
Cc: [email protected]
Subject: Re: [onap-discuss] [OOM] restart an entire POD ?

Hi Brian,

Those issues are tracked in JIRA already. Adding Mike that is looking at it (I 
think).

About your question, you cannot do this through K8S UI; at least, not that I’m 
aware of.
But using our scripts, you can delete and create a specific app.

For instance:
./oom/kubernetes/oneclick/deleteAll.sh -n onap -a aai <— will delete the whole 
AAI namespace (deployment and services) ./oom/kubernetes/oneclick/createAll.sh 
-n onap -a aai <— will create the whole AAI namespace (deployment and services)

I’m not sure this is what you’re after, but that’s how I do it when I need to 
bounce a whole application (e.g. all the containers of an app).

Alexis

> On Feb 7, 2018, at 9:34 AM, FREEMAN, BRIAN D <[email protected]> wrote:
> 
> Michael, Alexi,
> 
> I'm having race conditions when I use OOM in Azure where the health check 
> passes but distribution fails (MSO and AAI never get notified).
> 
> I restarted the SO front end POD and SO successfully picked up a model 
> distribution.
> 
> I tried to restart just the AAI Model loader but that didnt seem to 
> work so I need to restart all of AAI
> 
> I suspect that SO and AAI came up before DMaaP was up but cant confirm that.
> 
> Is there an easy / safe way to do restart an entire domain through the K8 
> portal ?
> 
> Feel free to point me at the right documentation on the wiki if I am just 
> missing that guidance.
> 
> Brian
> 
> _______________________________________________
> onap-discuss mailing list
> [email protected]
> https://lists.onap.org/mailman/listinfo/onap-discuss

_______________________________________________
onap-discuss mailing list
[email protected]
https://lists.onap.org/mailman/listinfo/onap-discuss
This message and the information contained herein is proprietary and 
confidential and subject to the Amdocs policy statement,

you may review at https://www.amdocs.com/about/email-disclaimer 
<https://www.amdocs.com/about/email-disclaimer>
_______________________________________________
onap-discuss mailing list
[email protected]
https://lists.onap.org/mailman/listinfo/onap-discuss

Reply via email to