The steps needed to upgrade the undercloud were
specified in two different files. This commit
updates the steps to cover O to P from commit
496223 in the 'Undercloud Installation' section
and reuses that doc file for the 'Upgrading to
a Next Major Release'.
As an addition, a new class 'otop' has been
added to include specific documentation sections
when upgrading from Ocata to Pike.
Change-Id: I9097ca11a95f94c5eaf580fb8a82b3173d606a5d
Update the undercloud upgrade for O to P
Update the controller and converge step command
with the right env file
Also adds a new section for the Pike minor update docs
Co-Authored-By: Marios Andreou <marios@redhat.com>
Change-Id: I7f7f49d7450f2b158d29b2d1b37dd3bcabfeb5f8
The install documentation is currently missing the pike repository
information. This change adds stable/pike into the repository
documentation.
Change-Id: Ia34da03343d726d4796013a816bd5ed53e9d3edf
Closes-Bug: #1723265
We need a short howto to explain to folks how to test code fixes that
affect the container itself.
Co-Authored-By: Ian Pilcher <ipilcher@redhat.com>
Change-Id: I0ac7901914bb6340e1552b19d16684981ce6cd05
We currently do the following to add the 'stack' user on the undercloud
to the 'docker' group so it can access the docker daemon (when the local
docker registry option is enabled):
...
if str2bool(hiera('enable_docker_registry', true)) {
group { 'docker':
ensure => 'present',
}
user { 'docker_user':
name => hiera('tripleo_install_user'),
groups => 'docker',
notify => Service['docker'],
}
include ::tripleo::profile::base::docker_registry
}
...
For this to work the stack user needs to logoff and logon again.
Closes-Bug: #1722498
Change-Id: Ia1f76f61828c119667879c41f1ad5d86b8901adb
This corrects the second example showing usage of DeployedServerPortMap,
where the subnets key was accidentally omitted.
Change-Id: I66562468f8c118eea634d931db6157b4b45b5c4b
Due to missing/superflous spaces, some of the content in the
guide is displayed as blockquote without being really quoted
content. This change adds/removes spaces to remove ALL the
generated HTML blockquotes.
Change-Id: I25b0d9fa64cd474a844b5f3e6c126395a4e80f2c
The undercloud update docs were changed to reflect the new process
that started in Pike, but the major upgrade docs were not. While the
old process will work for now, there are plans to make changes that
will rely on the new process so we need to make sure to document it.
See 1e731bd201
bp undercloud-upgrade
Change-Id: I8704b197471f5e53377a7a5f18cdcc43563c54cb
undercloud_hostname is a less error-prone method of configuring all
the undercloud hostname settings correctly. We probably want to
make it a mandatory option and avoid hostname-related bugs, but for
the moment let's at least document it as the preferred way to handle
hostname configuration.
Note that this option has existed since Mitaka, so it applies to
all of our currently supported branches.
Change-Id: I2506a44abac999d4b0450c402fe40cab89410cea
We haven't actually recommended installing anything from source for
quite a while and it probably doesn't work anymore anyway. Let's
get rid of this admonition and clean things up a little.
Change-Id: I005bc042f4d1ff855a90710599d796a9645b71f1
This is an attempt to document how to integrate a new container
or extend an existing one and include it in a TripleO deployment.
Change-Id: I10109341ad547645611337c582db1b63a349c5ae
Co-Authored-By: Dan Prince <dprince@redhat.com>
Uses better (shorter) examples to distinguish better which one is
to be used for what.
Change-Id: I51341fb3aa72106e48241a97c89bb3a7bc1d6fb8
Depends-On: I145feb0208f135da7c71694ebcecd937244d66b1
Related-Bug: #1719919
Some docs examples refer to the now deprecated NovaCompute* parameters,
so update to the new Compute* convention which is consistent with other
roles, and note the correct steps for map parameters (which are merged so
we don't want any old values left in the environment).
Change-Id: I13c75d06c482aa87bac4bab9347dd2ae643af44f
NeutronNetworkVLANRanges should be set too when storage network
uses VLAN otherwise neutron complains about wrong network type
when creating provider network.
It may be neccessary to shrink StorageAllocationPools to make
sure there are some IPs left for tenants.
Change-Id: I10b7945f12f9dc4180777c69221a4f57c6185719
Bring these docs up to date with the current state of containerization. Also
remove warnings as containers are now fully supported.
Change-Id: I1d53195b586f25f9fb6acd9aac0cc42b3857deaa
Co-Authored-By: Dan Prince <dprince@redhat.com>
Since we are now executing tempest in our gates, it is good to have
the steps required to enable and customize notifications documented
Change-Id: Ie5703eb43a9a7676aba711482797e8725d0f9878
Due to missing spaces, some of the content in the installation
guide is displayed as blockquote without being really quoted
content. This change adds the needed spaces to remove the
generated HTML blockquotes.
The repositories.txt file gets included at different locations
and that file contained the formatting spaces which were needed
from the location which includes that file. It's most of the time
better to do the indentation at the ".. include" level and leave
the file which gets included *without* indentation. Because here,
the indentation in the included file worked for some includes but
not for all, because the content got used at different indentation
levels. That made it necessary to touch the includes in
basic_deployment and post_deployment too.
Follow up patches will deal with the other install guide sections
more in details.
Change-Id: Ie2e7fea6d92f7742d23e22452e394b8ff40975eb
The existing list skips things like openstack-tripleo-validations and we
should just be including all the openstack-tripleo-* packages
Change-Id: I35e9540db61afa923aff7ab13bb95432d9fa529e
Related-Bug: #1715625
Entries should never be removed from the predictable ip lists, even
if a node is deleted. This is because the index into the list is
based on the Heat resource index, which doesn't change when a node
is deleted. This has come up a number of times in real-world
deployments so we should make this behavior clear in the docs.
Change-Id: I455a1e15bfc3cef07f6fd3c8462428985f12c0da
It is possible to limit the container images that are pulled from the
registry by specifying the environments and role files to the prepare
command.
Change-Id: Id815e865d68e47d05a408b987c638ec917ca4dfe
Related-Bug: #1710992
When repeating an overcloud deployment which reuses the same nodes
or disks of a previous deplyment, it is necessary to clean the
Ceph OSD disks for the activation to complete. Both puppet-ceph and
ceph-ansible have the same requirement.
This change documents use of Ironic automated_clean functionality
to automate the disks cleanup.
Change-Id: I08ba35c59d3a6aaed59a1526757a4fa513acb190
Related-Bug: #1714242
Due to the way that the new documentation layout is set up. The TLS
everywhere title wasn't showing up. This fixes that.
Change-Id: I30795abdf231785b4df715dc53b42840c92d1a20
The overcloud might need to trust the undercloud's CA certificate. So
this was noted explicitly in the docs.
Change-Id: Id3651ae48a72ce9e1a6e5bafd209825f4fecb8c4
Jose Luis Franco Arza