One gotcha with using "profile set" as the first statement is the user gets
an off-putting message

./minishift profile set istio-demo
oc cli context could not changed for 'istio-demo'. Make sure the profile is
in running state or restart if the problem persists.

the problem with this message is that I have several more commands to run
before I run "start", so stop complaining to me that I have not yet ran
start.



On Sun, Nov 5, 2017 at 7:26 PM, Gerard Braad <[email protected]> wrote:

> Hi,
>
>
> On Sun, Nov 5, 2017 at 4:39 PM, Burr Sutter <[email protected]> wrote:
> > but one syntax requires no "--" and the other has  "--" which I find to
> be
> > odd since I am using the commands back to back.
>
> One is a command to set the active profile for subsequent commands,
> while the other is a flag to perform a single action/command outside
> of the currently chosen or active profile.
>
> On Sun, Nov 5, 2017 at 11:35 PM, Burr Sutter <[email protected]> wrote:
> > I was able to switch profiles live today.
>
> Good to hear! There is actually not much to it, we do still have some
> kinks to iron out.
> Feedback like yours will help with, as some can be as easy as better
> feedback during start, documentation, etc
>
> On Mon, Nov 6, 2017 at 6:33 AM, Burr Sutter <[email protected]> wrote:
> > But for subsequent restarts, running that same sequence yields lots of
> > messages that look bad to the end-user.
>
> At the moment we provide a verbose feedback about the startup of
> Minishift, as we experience many reports related to missing settings
> or something related to being on a different distro, or just sometimes
> weird behaviour...
>
> We will try to reduce the output after we are a little bit more
> stable, and only report things when we really see a failure. This is
> pretty much the same as what happened for the output of `oc cluster
> up`, as they also dropped the verbosity after some time.
>
> > The big win on profiles is the ability to start them and stop them
> often, on the fly, in front of a live audience.
>
> As long as the IP address doesn't chane... as we still have an issue
> with how certificates are generated by `oc cluster up`. As long as
> there is no way to regenerate the certificates for the new IP address
> assigned to the VM, the startup would fail. We can work around this by
> forcibly assigning an address on restart, but this will only work with
> using the same address when the certificate was generated.
>
> I hope this can be solved..
>
>
> Gerard
>
_______________________________________________
Devtools mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/devtools

Reply via email to