Re: Tutorial for Minishift
by Jorge Morales Pou
What are you looking for? A tutorial on how to use minishift in an
affective way or a tutorial of developing in openshift? If the former, the
example itself is not so important, but rather the workflow and commands.
Jorge
El 13 feb. 2018 10:13, "Andreas Gajdosik" <agajdosi(a)redhat.com> escribió:
Big +1.
> From: "Budh Ram Gurung" <bgurung(a)redhat.com>
> To: "Lalatendu Mohanty" <lmohanty(a)redhat.com>
> Cc: "Minishift Mailing List" <minishift(a)lists.minishift.io>, "Gerard
Braad" <gbraad(a)redhat.com>, "Anjan Nath"
> <kaludios(a)gmail.com>, "Robin Owen" <robin.owen(a)redhat.com>
> Sent: Tuesday, February 13, 2018 9:43:18 AM
> Subject: [minishift] Re: Tutorial for Minishift
>
> + Robin
>
> On Tue, Feb 13, 2018 at 2:09 PM, Lalatendu Mohanty < lmohanty(a)redhat.com >
> wrote:
>
>
>
>
>
> On Tue, Feb 13, 2018 at 2:03 PM, Gerard Braad < gbraad(a)redhat.com > wrote:
>
>
> Hi,
>
>
> This is pretty much part of the on-going discussion of "articles"
> around Minishift. I made a call about topics and goals we wanted to
> show, but so far this "issue" got closed by staleness. Seems to make
> this happen it needs to be lean...
>
> So, if we want to do this, I would again suggest to describe the same
> effort as the basic feature does, but in a step by step fashion (oh
> wait, the feature file already is ;-), because it is a proven to work
> scenario. And from there we can expand on it.
>
> also, what does Robin think about the inclusion in our documentation?
>
> What do you think?
>
>
>
>
>
> Or should it be somewhere else?
>
>
> I would suggest to keep it part of the official documentation. We have
been
> seeing issues where users not know that they can use "config set" commands
> to persist configuration etc. So making part of the documentation and then
> pointing users towards it will help educating users about it.
>
> -Lala
>
Agree, seems fine to have one or few tutorial as part of the documentation.
As user I appreciate when getting started, tutorial and more precise
documentation
are in the same place. One can be then more sure that he is not reading
something
outdated. However we can also have other tutorials outside of the Minishift
documentation,
would be nice to have more of them. (Around Christmas I have played a
little bit with
openshift/golang-ex, maybe I could contribute with it later.)
We can also write a test around the tutorial strictly following commands
used in it to
make sure these few tutorials are 'just working'.
- Andreas
>
>
> regards,
>
> Gerard
> _______________________________________________
> Minishift Mailing List mailing list -- minishift(a)lists.minishift.io
> To unsubscribe send an email to minishift-leave(a)lists.minishift.io
>
>
> _______________________________________________
> Minishift Mailing List mailing list -- minishift(a)lists.minishift.io
> To unsubscribe send an email to minishift-leave(a)lists.minishift.io
>
>
> Regards,
> Budh Ram Gurung
>
>
> _______________________________________________
> Minishift Mailing List mailing list -- minishift(a)lists.minishift.io
> To unsubscribe send an email to minishift-leave(a)lists.minishift.io
>
_______________________________________________
Minishift Mailing List mailing list -- minishift(a)lists.minishift.io
To unsubscribe send an email to minishift-leave(a)lists.minishift.io
4 years, 11 months
Fwd: [aos-devel] Summary of 3.10 oc cluster up changes
by Adam Kaplan
FYI folks - I spent my "Hack Day" in the Red Hat Raleigh office effectively
researching the impact of these changes on minishift. There's quite a lot,
and it wasn't clear if all of these were captured on the Github roadmap:
1: I *think* running each component separately may not be an issue, as
minishift start is just a shell around cluster up. Obviously needs to be
tested.
2,3,9: minishift start will need a separate strategy to handle v3.10's
flags, as these flags are supported by minishift and will *not* be
supported in 3.10.
4: --use-existing-config is hard-coded into the cluster up command. This
needs to be removed.
5: May only break a few unit tests.
6: New feature provided by oc client
7,8: Referenced in the docs, but always as experimental flags to pass into
`minishift start`. Seems like we will push users to `oc cluster add` once
these are working.
Please let me know if there are existing Trello cards/Github issues for
these items - if not I'd be happy to write these up.
Thanks,
Adam
---------- Forwarded message ----------
From: David Eads <deads(a)redhat.com>
Date: Thu, Apr 5, 2018 at 11:14 AM
Subject: [aos-devel] Summary of 3.10 oc cluster up changes
To: aos-devel <aos-devel(a)redhat.com>
There are numerous changes related to `oc cluster up` in 3.10. A few of
the highlights:
1. `oc cluster up` no longer runs an all in one server. It runs a
kubelet, then runs static pods for etcd, an openshift-flavored
kube-apiserver, kube-controller-manager, and kube-scheduler. Then it runs
workload managed pods for a kube-dns, kube-proxy, openshift-apiserver, and
openshift-controller-manager.
2. `oc cluster up` uses a `--base-dir` which defaults to
`CWD/openshift.local.clusterup` to hold all of its config, etcd data,
volumes, and installation logs.
3. `--host-config-dir`, `--host-volumes-dir`, `--host-data-dir`, and
`host-pv-dir` have been removed in favor of the opinionated `--base-dir`
4. `oc cluster up` no longer uses `--use-existing-config`. Like the
existing `openshift start` behavior, it will use the `--base-dir` if it
exists. If you want to clear your data and start fresh, you'll need to
delete your config data and re-execute. If you want to re-use your data,
you don't need to do anything.
5. `--create-machine` and `--docker-machine` was removed. If you're
using a mac, you'll need to configure docker ahead of time.
6. `oc cluster add <component>` where component is something like
"service-catalog" and "template-service-broker" can be used to add a
component to your already existing `oc cluster up` cluster. This adds
flexibility and allows you to easily add features as you want them.
7. `--service-catalog` has been replaced by `oc cluster up --enable
service-catalog,template-service-broker`. Since `oc cluster add` now
exists, you can enable any component you can `oc cluster add` by adding it
to that list.
8. `--logging` and `--metrics` have been removed. They were
experimental and have not worked for quite some time. We recognize it is
important functionality, but since it was already non-functional, it was
experimental, `oc cluster add` exists, and the deployment topology changed,
they need some non-trivial rework. When they come back in, they should be
part of `oc cluster add`.
9. `--version` was removed. We didn't previously support version skew.
Instead, `oc cluster up` uses the rolling tag (v3.10.0) instead. If you
need to specify a particular tag, you can use `--tag`, but be aware that it
is not a version. We do not parse it or interpret it.
We are likely to transition our other existing `oc cluster up` components
to be available using `oc cluster add`. If you have to report a bug in
cluster up, please include the content from `{--base-dir}/logs` to ease
debugging.
--
ADAM KAPLAN
SENIOR SOFTWARE ENGINEER - OPENSHIFT
Red Hat <https://www.redhat.com/>
100 E Davie St Raleigh, NC 27601 USA
adam.kaplan(a)redhat.com T: +1-919-754-4843 IM: adambkaplan
<https://red.ht/sig>
4 years, 11 months
Proposal to add new label for size of work required
by Anjan Nath
Hi All,
As the subject says, I feel adding labels like (size/big)
(size/small), etc, to the
issues would be helpful to outside contributors, for deciding if they
want to work
on it, as well as we can sync our sprint planning task size to github
issues that
way. It will also be helpful in scenarios where we do some spike around an issue
and we can then mark on the issue how much work would be required.
What do you guys think?
Thanks,
Anjan
4 years, 11 months
Release announcement for Minishift 1.16.0
by Lalatendu Mohanty
Hi All,
The Minishift team is pleased to announce the release of Minishift 1.16.0.
Minishift [1] is a command-line tool that provisions and manages a
single-node
OpenShift cluster. You can run Minishift on GNU/Linux,Microsoft Windows or
macOS.
In version 1.16.0, we added new features, enhancements and bug fixes
including:
- Local DNS server for offline use (Experimental feature)
Now routes can be accessed offline without dependency on nip.io dns
resolver. See documentation [4].
- service-catalog, logging and metrics flags needs to be used as part of
--extra-clusterup-flags
E.g. minishift start --extra-clusterup-flags "--service-catalog” .
Note: These flags do not work reliably and also removed from next version
of “oc cluster up” i.e. (v 3.10.0).
You can find a detailed list of issues in the release notes[2].
For information about getting started, using, and developing Minishift,
see the documentation [5].
Please try the new release.
The easiest upgrade path is by simply running:
$ minishift update
If you are on v1.15.0 or earlier versions, ‘minishift update’ command will
fail to update to v1.16.0 because of the issue [3]. So you need to download
v1.16.0 from minishift github releases page [2] manually
to use the released binary.
We are looking forward to your feedback.
The Minishift community [6] hangs out at #minishift channel on Freenode
and it is the perfect place to discuss anything Minishift related.
You can sign up for our mailing list [7] to receive development updates,
ask questions about Minishift, suggest ideas and talk to the community.
[1] https://github.com/minishift/minishift
[2] https://github.com/minishift/minishift/releases/tag/v1.16.0
[3] https://github.com/minishift/minishift/issues/2189
[4] <https://github.com/minishift/minishift/releases/tag/v1.13.0>
https://docs.openshift.org/latest/minishift/using/experimental-features.h...
[5] https://docs.openshift.org/latest/minishift/index.html
[6] https://github.com/minishift/minishift#community
[7] https://lists.minishift.io/admin/lists/minishift.lists.minishift.io
4 years, 11 months
Minishift Tomcat ClassPath Issue.
by mhpatel＠ibaset.com
Hi,
I am new to minishift and trying to deploy existing (Tomcat web application) Docker Image to minishift. I have created shared storage and create config maps where I have added log4j.properties file. I want to override my existing application log4j.properties with external one.
In existing container tomcat context resources xml file at (tomcat/conf/Catalina/localhost) has following content
<Resources className="org.apache.catalina.webresources.StandardRoot">
<PreResources
className="org.apache.catalina.webresources.DirResourceSet"
base="/opt/config"
webAppMount="/WEB-INF/classes" />
</Resources>
So basically I am indicating tomcat to add /opt/config/ directory in classpath and also adding config map at the same location (volume binding) /opt/config but somewhat how it's not loading external log4j.properties from volume mapped location and loading default log4j.properties only.
Please suggest what's is going wrong.
Thanks
4 years, 12 months
