.. This work is licensed under a Creative Commons Attribution 4.0 International License.
.. http://creativecommons.org/licenses/by/4.0
.. Copyright 2018 Amdocs, Bell Canada

.. Links
.. _Curated applications for Kubernetes: https://github.com/kubernetes/charts
.. _Services: https://kubernetes.io/docs/concepts/services-networking/service/
.. _ReplicaSet: https://kubernetes.io/docs/concepts/workloads/controllers/replicaset/
.. _StatefulSet: https://kubernetes.io/docs/concepts/workloads/controllers/statefulset/
.. _Helm Documentation: https://docs.helm.sh/helm/
.. _Helm: https://docs.helm.sh/
.. _Kubernetes: https://Kubernetes.io/
.. _Kubernetes LoadBalancer: https://kubernetes.io/docs/concepts/services-networking/service/#type-loadbalancer
.. _user-guide-label:

OOM User Guide
##############

The ONAP Operations Manager (OOM) provide the ability to manage the entire
life-cycle of an ONAP installation, from the initial deployment to final
decommissioning. This guide provides instructions for users of ONAP to
use the Kubernetes_/Helm_ system as a complete ONAP management system.

This guide provides many examples of Helm command line operations.  For a
complete description of these commands please refer to the `Helm
Documentation`_.

.. figure:: oomLogoV2-medium.png
   :align: right

The following sections describe the life-cycle operations:

- Deploy_ - with built-in component dependency management
- Configure_ - unified configuration across all ONAP components
- Monitor_ - real-time health monitoring feeding to a Consul UI and Kubernetes
- Heal_- failed ONAP containers are recreated automatically
- Scale_ - cluster ONAP services to enable seamless scaling
- Upgrade_ - change-out containers or configuration with little or no service impact
- Delete_ - cleanup individual containers or entire deployments

.. figure:: oomLogoV2-Deploy.png
   :align: right

Deploy
======

The OOM team with assistance from the ONAP project teams, have built a
comprehensive set of Helm charts, yaml files very similar to TOSCA files, that
describe the composition of each of the ONAP components and the relationship
within and between components. Using this model Helm is able to deploy all of
ONAP with a few simple commands.

Pre-requisites
--------------
Your environment must have both the Kubernetes `kubectl` and Helm setup as a
one time activity.

Install Kubectl
~~~~~~~~~~~~~~~
Enter the following to install kubectl (on Ubuntu, there are slight differences
on other O/Ss), the Kubernetes command line interface used to manage a
Kubernetes cluster::

  > curl -LO https://storage.googleapis.com/kubernetes-release/release/v1.8.10/bin/linux/amd64/kubectl
  > chmod +x ./kubectl
  > sudo mv ./kubectl /usr/local/bin/kubectl
  > mkdir ~/.kube

Paste kubectl config from Rancher (see the :ref:`cloud-setup-guide-label` for
alternative Kubernetes environment setups) into the `~/.kube/config` file.

Verify that the Kubernetes config is correct::

  > kubectl get pods --all-namespaces

At this point you should see six Kubernetes pods running.

Install Helm
~~~~~~~~~~~~
Helm is used by OOM for package and configuration management. To install Helm,
enter the following::

  > wget http://storage.googleapis.com/kubernetes-helm/helm-v2.9.1-linux-amd64.tar.gz
  > tar -zxvf helm-v2.9.1-linux-amd64.tar.gz
  > sudo mv linux-amd64/helm /usr/local/bin/helm

Verify the Helm version with::

  > helm version

Install the Helm Tiller application and initialize with::

  > helm init

Install the Helm Repo
---------------------
Once kubectl and Helm are setup, one needs to setup a local Helm server to
server up the ONAP charts::

  > helm install osn/onap

.. note::
  The osn repo is not currently available so creation of a local repository is
  required.

Helm is able to use charts served up from a repository and comes setup with a
default CNCF provided `Curated applications for Kubernetes`_ repository called
stable which should be removed to avoid confusion::

  > helm repo remove stable

.. To setup the Open Source Networking Nexus repository for helm enter::
..  > helm repo add osn 'https://nexus3.onap.org:10001/helm/helm-repo-in-nexus/master/'

To prepare your system for an installation of ONAP, you'll need to::

  > git clone -b casablanca http://gerrit.onap.org/r/oom
  > cd oom/kubernetes


To setup a local Helm server to server up the ONAP charts::

  > helm init
  > helm serve &

Note the port number that is listed and use it in the Helm repo add as
follows::

  > helm repo add local http://127.0.0.1:8879

To get a list of all of the available Helm chart repositories::

  > helm repo list
  NAME   URL
  local  http://127.0.0.1:8879

Then build your local Helm repository::

  > make all

The Helm search command reads through all of the repositories configured on the
system, and looks for matches::

  > helm search -l
  NAME                    VERSION    DESCRIPTION
  local/appc              2.0.0      Application Controller
  local/clamp             2.0.0      ONAP Clamp
  local/common            2.0.0      Common templates for inclusion in other charts
  local/onap              2.0.0      Open Network Automation Platform (ONAP)
  local/robot             2.0.0      A helm Chart for kubernetes-ONAP Robot
  local/so                2.0.0      ONAP Service Orchestrator

In any case, setup of the Helm repository is a one time activity.

Next, install Helm Plugins required to deploy the ONAP Casablanca release::

  > cp -R helm/plugins/ ~/.helm

Once the repo is setup, installation of ONAP can be done with a single
command::

  > helm deploy development local/onap --namespace onap

This will install ONAP from a local repository in a 'development' Helm release.
As described below, to override the default configuration values provided by
OOM, an environment file can be provided on the command line as follows::

  > helm deploy development local/onap --namespace onap -f overrides.yaml

To get a summary of the status of all of the pods (containers) running in your
deployment::

  > kubectl get pods --all-namespaces -o=wide

.. note::
  The Kubernetes namespace concept allows for multiple instances of a component
  (such as all of ONAP) to co-exist with other components in the same
  Kubernetes cluster by isolating them entirely.  Namespaces share only the
  hosts that form the cluster thus providing isolation between production and
  development systems as an example.  The OOM deployment of ONAP in Beijing is
  now done within a single Kubernetes namespace where in Amsterdam a namespace
  was created for each of the ONAP components.

.. note::
  The Helm `--name` option refers to a release name and not a Kubernetes namespace.


To install a specific version of a single ONAP component (`so` in this example)
with the given release name enter::

  > helm deploy so onap/so --version 3.0.1

To display details of a specific resource or group of resources type::

  > kubectl describe pod so-1071802958-6twbl

where the pod identifier refers to the auto-generated pod identifier.

.. figure:: oomLogoV2-Configure.png
   :align: right

Configure
=========

Each project within ONAP has its own configuration data generally consisting
of: environment variables, configuration files, and database initial values.
Many technologies are used across the projects resulting in significant
operational complexity and an inability to apply global parameters across the
entire ONAP deployment. OOM solves this problem by introducing a common
configuration technology, Helm charts, that provide a hierarchical
configuration with the ability to override values with higher
level charts or command line options.

The structure of the configuration of ONAP is shown in the following diagram.
Note that key/value pairs of a parent will always take precedence over those
of a child. Also note that values set on the command line have the highest
precedence of all.

.. graphviz::

   digraph config {
      {
         node     [shape=folder]
         oValues  [label="values.yaml"]
         demo     [label="onap-demo.yaml"]
         prod     [label="onap-production.yaml"]
         oReq     [label="requirements.yaml"]
         soValues [label="values.yaml"]
         soReq    [label="requirements.yaml"]
         mdValues [label="values.yaml"]
      }
      {
         oResources  [label="resources"]
      }
      onap -> oResources
      onap -> oValues
      oResources -> environments
      oResources -> oReq
      oReq -> so
      environments -> demo
      environments -> prod
      so -> soValues
      so -> soReq
      so -> charts
      charts -> mariadb
      mariadb -> mdValues

   }

The top level onap/values.yaml file contains the values required to be set
before deploying ONAP.  Here is the contents of this file:

.. include:: onap_values.yaml
   :code: yaml

One may wish to create a value file that is specific to a given deployment such
that it can be differentiated from other deployments.  For example, a
onap-development.yaml file may create a minimal environment for development
while onap-production.yaml might describe a production deployment that operates
independently of the developer version.

For example, if the production OpenStack instance was different from a
developer's instance, the onap-production.yaml file may contain a different
value for the vnfDeployment/openstack/oam_network_cidr key as shown below.

.. code-block:: yaml

  nsPrefix: onap
  nodePortPrefix: 302
  apps: consul msb mso message-router sdnc vid robot portal policy appc aai
  sdc dcaegen2 log cli multicloud clamp vnfsdk aaf kube2msb
  dataRootDir: /dockerdata-nfs

  # docker repositories
  repository:
    onap: nexus3.onap.org:10001
    oom: oomk8s
    aai: aaionap
    filebeat: docker.elastic.co

  image:
    pullPolicy: Never

  # vnf deployment environment
  vnfDeployment:
    openstack:
      ubuntu_14_image: "Ubuntu_14.04.5_LTS"
      public_net_id: "e8f51956-00dd-4425-af36-045716781ffc"
      oam_network_id: "d4769dfb-c9e4-4f72-b3d6-1d18f4ac4ee6"
      oam_subnet_id: "191f7580-acf6-4c2b-8ec0-ba7d99b3bc4e"
      oam_network_cidr: "192.168.30.0/24"
  <...>


To deploy ONAP with this environment file, enter::

  > helm deploy local/onap -n casablanca -f environments/onap-production.yaml

.. include:: environments_onap_demo.yaml
   :code: yaml

When deploying all of ONAP a requirements.yaml file control which and what
version of the ONAP components are included.  Here is an excerpt of this
file:

.. code-block:: yaml

  # Referencing a named repo called 'local'.
  # Can add this repo by running commands like:
  # > helm serve
  # > helm repo add local http://127.0.0.1:8879
  dependencies:
  <...>
    - name: so
      version: ~2.0.0
      repository: '@local'
      condition: so.enabled
  <...>

The ~ operator in the `so` version value indicates that the latest "2.X.X"
version of `so` shall be used thus allowing the chart to allow for minor
upgrades that don't impact the so API; hence, version 2.0.1 will be installed
in this case.

The onap/resources/environment/onap-dev.yaml (see the excerpt below) enables
for fine grained control on what components are included as part of this
deployment. By changing this `so` line to `enabled: false` the `so` component
will not be deployed.  If this change is part of an upgrade the existing `so`
component will be shut down. Other `so` parameters and even `so` child values
can be modified, for example the `so`'s `liveness` probe could be disabled
(which is not recommended as this change would disable auto-healing of `so`).

.. code-block:: yaml

  #################################################################
  # Global configuration overrides.
  #
  # These overrides will affect all helm charts (ie. applications)
  # that are listed below and are 'enabled'.
  #################################################################
  global:
  <...>

  #################################################################
  # Enable/disable and configure helm charts (ie. applications)
  # to customize the ONAP deployment.
  #################################################################
  aaf:
    enabled: false
  <...>
  so: # Service Orchestrator
    enabled: true

    replicaCount: 1

    liveness:
      # necessary to disable liveness probe when setting breakpoints
      # in debugger so K8s doesn't restart unresponsive container
      enabled: true

  <...>

Accessing the ONAP Portal using OOM and a Kubernetes Cluster
------------------------------------------------------------

The ONAP deployment created by OOM operates in a private IP network that isn't
publicly accessible (i.e. Openstack VMs with private internal network) which
blocks access to the ONAP Portal. To enable direct access to this Portal from a
user's own environment (a laptop etc.) the portal application's port 8989 is
exposed through a `Kubernetes LoadBalancer`_ object.

Typically, to be able to access the Kubernetes nodes publicly a public address
is assigned. In Openstack this is a floating IP address.

When the `portal-app` chart is deployed a Kubernetes service is created that
instantiates a load balancer.  The LB chooses the private interface of one of
the nodes as in the example below (10.0.0.4 is private to the K8s cluster only).
Then to be able to access the portal on port 8989 from outside the K8s &
Openstack environment, the user needs to assign/get the floating IP address that
corresponds to the private IP as follows::

  > kubectl -n onap get services|grep "portal-app"
  portal-app  LoadBalancer   10.43.142.201   10.0.0.4   8989:30215/TCP,8006:30213/TCP,8010:30214/TCP   1d   app=portal-app,release=dev


In this example, use the 10.0.0.4 private address as a key find the
corresponding public address which in this example is 10.12.6.155. If you're
using OpenStack you'll do the lookup with the horizon GUI or the Openstack CLI
for your tenant (openstack server list).  That IP is then used in your
`/etc/hosts` to map the fixed DNS aliases required by the ONAP Portal as shown
below::

  10.12.6.155 portal.api.simpledemo.onap.org
  10.12.6.155 vid.api.simpledemo.onap.org
  10.12.6.155 sdc.api.fe.simpledemo.onap.org
  10.12.6.155 sdc.workflow.plugin.simpledemo.onap.org
  10.12.6.155 sdc.dcae.plugin.simpledemo.onap.org
  10.12.6.155 portal-sdk.simpledemo.onap.org
  10.12.6.155 policy.api.simpledemo.onap.org
  10.12.6.155 aai.api.sparky.simpledemo.onap.org
  10.12.6.155 cli.api.simpledemo.onap.org
  10.12.6.155 msb.api.discovery.simpledemo.onap.org
  10.12.6.155 msb.api.simpledemo.onap.org
  10.12.6.155 clamp.api.simpledemo.onap.org
  10.12.6.155 so.api.simpledemo.onap.org

Ensure you've disabled any proxy settings the browser you are using to access
the portal and then simply access now the new ssl-encrypted URL:
https://portal.api.simpledemo.onap.org:30225/ONAPPORTAL/login.htm

.. note::
  Using the HTTPS based Portal URL the Browser needs to be configured to accept
  unsecure credentials.
  Additionally when opening an Application inside the Portal, the Browser
  might block the content, which requires to disable the blocking and reloading
  of the page

.. note::
  Besides the ONAP Portal the Components can deliver additional user interfaces,
  please check the Component specific documentation.

.. note::

   | Alternatives Considered:

   -  Kubernetes port forwarding was considered but discarded as it would require
      the end user to run a script that opens up port forwarding tunnels to each of
      the pods that provides a portal application widget.

   -  Reverting to a VNC server similar to what was deployed in the Amsterdam
      release was also considered but there were many issues with resolution, lack
      of volume mount, /etc/hosts dynamic update, file upload that were a tall order
      to solve in time for the Beijing release.

   Observations:

   -  If you are not using floating IPs in your Kubernetes deployment and directly attaching
      a public IP address (i.e. by using your public provider network) to your K8S Node
      VMs' network interface, then the output of 'kubectl -n onap get services | grep "portal-app"'
      will show your public IP instead of the private network's IP. Therefore,
      you can grab this public IP directly (as compared to trying to find the floating
      IP first) and map this IP in /etc/hosts.

.. figure:: oomLogoV2-Monitor.png
   :align: right

Monitor
=======

All highly available systems include at least one facility to monitor the
health of components within the system.  Such health monitors are often used as
inputs to distributed coordination systems (such as etcd, zookeeper, or consul)
and monitoring systems (such as nagios or zabbix). OOM provides two mechanisms
to monitor the real-time health of an ONAP deployment:

- a Consul GUI for a human operator or downstream monitoring systems and
  Kubernetes liveness probes that enable automatic healing of failed
  containers, and
- a set of liveness probes which feed into the Kubernetes manager which
  are described in the Heal section.

Within ONAP, Consul is the monitoring system of choice and deployed by OOM in
two parts:

- a three-way, centralized Consul server cluster is deployed as a highly
  available monitor of all of the ONAP components, and
- a number of Consul agents.

The Consul server provides a user interface that allows a user to graphically
view the current health status of all of the ONAP components for which agents
have been created - a sample from the ONAP Integration labs follows:

.. figure:: consulHealth.png
   :align: center

To see the real-time health of a deployment go to: http://<kubernetes IP>:30270/ui/
where a GUI much like the following will be found:


.. figure:: oomLogoV2-Heal.png
   :align: right

Heal
====

The ONAP deployment is defined by Helm charts as mentioned earlier.  These Helm
charts are also used to implement automatic recoverability of ONAP components
when individual components fail. Once ONAP is deployed, a "liveness" probe
starts checking the health of the components after a specified startup time.

Should a liveness probe indicate a failed container it will be terminated and a
replacement will be started in its place - containers are ephemeral. Should the
deployment specification indicate that there are one or more dependencies to
this container or component (for example a dependency on a database) the
dependency will be satisfied before the replacement container/component is
started. This mechanism ensures that, after a failure, all of the ONAP
components restart successfully.

To test healing, the following command can be used to delete a pod::

  > kubectl delete pod [pod name] -n [pod namespace]

One could then use the following command to monitor the pods and observe the
pod being terminated and the service being automatically healed with the
creation of a replacement pod::

  > kubectl get pods --all-namespaces -o=wide

.. figure:: oomLogoV2-Scale.png
   :align: right

Scale
=====

Many of the ONAP components are horizontally scalable which allows them to
adapt to expected offered load.  During the Beijing release scaling is static,
that is during deployment or upgrade a cluster size is defined and this cluster
will be maintained even in the presence of faults. The parameter that controls
the cluster size of a given component is found in the values.yaml file for that
component.  Here is an excerpt that shows this parameter:

.. code-block:: yaml

  # default number of instances
  replicaCount: 1

In order to change the size of a cluster, an operator could use a helm upgrade
(described in detail in the next section) as follows::

  > helm upgrade --set replicaCount=3 onap/so/mariadb

The ONAP components use Kubernetes provided facilities to build clustered,
highly available systems including: Services_ with load-balancers, ReplicaSet_,
and StatefulSet_.  Some of the open-source projects used by the ONAP components
directly support clustered configurations, for example ODL and MariaDB Galera.

The Kubernetes Services_ abstraction to provide a consistent access point for
each of the ONAP components, independent of the pod or container architecture
of that component.  For example, SDN-C uses OpenDaylight clustering with a
default cluster size of three but uses a Kubernetes service to and change the
number of pods in this abstract this cluster from the other ONAP components
such that the cluster could change size and this change is isolated from the
other ONAP components by the load-balancer implemented in the ODL service
abstraction.

A ReplicaSet_ is a construct that is used to describe the desired state of the
cluster.  For example 'replicas: 3' indicates to Kubernetes that a cluster of 3
instances is the desired state.  Should one of the members of the cluster fail,
a new member will be automatically started to replace it.

Some of the ONAP components many need a more deterministic deployment; for
example to enable intra-cluster communication. For these applications the
component can be deployed as a Kubernetes StatefulSet_ which will maintain a
persistent identifier for the pods and thus a stable network id for the pods.
For example: the pod names might be web-0, web-1, web-{N-1} for N 'web' pods
with corresponding DNS entries such that intra service communication is simple
even if the pods are physically distributed across multiple nodes. An example
of how these capabilities can be used is described in the Running Consul on
Kubernetes tutorial.

.. figure:: oomLogoV2-Upgrade.png
   :align: right

Upgrade
=======

Helm has built-in capabilities to enable the upgrade of pods without causing a
loss of the service being provided by that pod or pods (if configured as a
cluster).  As described in the OOM Developer's Guide, ONAP components provide
an abstracted 'service' end point with the pods or containers providing this
service hidden from other ONAP components by a load balancer. This capability
is used during upgrades to allow a pod with a new image to be added to the
service before removing the pod with the old image. This 'make before break'
capability ensures minimal downtime.

Prior to doing an upgrade, determine of the status of the deployed charts::

  > helm list
  NAME REVISION UPDATED                  STATUS    CHART     NAMESPACE
  so   1        Mon Feb 5 10:05:22 2018  DEPLOYED  so-2.0.1  default

When upgrading a cluster a parameter controls the minimum size of the cluster
during the upgrade while another parameter controls the maximum number of nodes
in the cluster.  For example, SNDC configured as a 3-way ODL cluster might
require that during the upgrade no fewer than 2 pods are available at all times
to provide service while no more than 5 pods are ever deployed across the two
versions at any one time to avoid depleting the cluster of resources. In this
scenario, the SDNC cluster would start with 3 old pods then Kubernetes may add
a new pod (3 old, 1 new), delete one old (2 old, 1 new), add two new pods (2
old, 3 new) and finally delete the 2 old pods (3 new).  During this sequence
the constraints of the minimum of two pods and maximum of five would be
maintained while providing service the whole time.

Initiation of an upgrade is triggered by changes in the Helm charts.  For
example, if the image specified for one of the pods in the SDNC deployment
specification were to change (i.e. point to a new Docker image in the nexus3
repository - commonly through the change of a deployment variable), the
sequence of events described in the previous paragraph would be initiated.

For example, to upgrade a container by changing configuration, specifically an
environment value::

  > helm deploy casablanca onap/so --version 2.0.1 --set enableDebug=true

Issuing this command will result in the appropriate container being stopped by
Kubernetes and replaced with a new container with the new environment value.

To upgrade a component to a new version with a new configuration file enter::

  > helm deploy casablanca onap/so --version 2.0.2 -f environments/demo.yaml

To fetch release history enter::

  > helm history so
  REVISION UPDATED                  STATUS     CHART     DESCRIPTION
  1        Mon Feb 5 10:05:22 2018  SUPERSEDED so-2.0.1  Install complete
  2        Mon Feb 5 10:10:55 2018  DEPLOYED   so-2.0.2  Upgrade complete

Unfortunately, not all upgrades are successful.  In recognition of this the
lineup of pods within an ONAP deployment is tagged such that an administrator
may force the ONAP deployment back to the previously tagged configuration or to
a specific configuration, say to jump back two steps if an incompatibility
between two ONAP components is discovered after the two individual upgrades
succeeded.

This rollback functionality gives the administrator confidence that in the
unfortunate circumstance of a failed upgrade the system can be rapidly brought
back to a known good state.  This process of rolling upgrades while under
service is illustrated in this short YouTube video showing a Zero Downtime
Upgrade of a web application while under a 10 million transaction per second
load.

For example, to roll-back back to previous system revision enter::

  > helm rollback so 1

  > helm history so
  REVISION UPDATED                  STATUS     CHART     DESCRIPTION
  1        Mon Feb 5 10:05:22 2018  SUPERSEDED so-2.0.1  Install complete
  2        Mon Feb 5 10:10:55 2018  SUPERSEDED so-2.0.2  Upgrade complete
  3        Mon Feb 5 10:14:32 2018  DEPLOYED   so-2.0.1  Rollback to 1

.. note::

  The description field can be overridden to document actions taken or include
  tracking numbers.

Many of the ONAP components contain their own databases which are used to
record configuration or state information.  The schemas of these databases may
change from version to version in such a way that data stored within the
database needs to be migrated between versions. If such a migration script is
available it can be invoked during the upgrade (or rollback) by Container
Lifecycle Hooks. Two such hooks are available, PostStart and PreStop, which
containers can access by registering a handler against one or both. Note that
it is the responsibility of the ONAP component owners to implement the hook
handlers - which could be a shell script or a call to a specific container HTTP
endpoint - following the guidelines listed on the Kubernetes site. Lifecycle
hooks are not restricted to database migration or even upgrades but can be used
anywhere specific operations need to be taken during lifecycle operations.

OOM uses Helm K8S package manager to deploy ONAP components. Each component is
arranged in a packaging format called a chart - a collection of files that
describe a set of k8s resources. Helm allows for rolling upgrades of the ONAP
component deployed. To upgrade a component Helm release you will need an
updated Helm chart. The chart might have modified, deleted or added values,
deployment yamls, and more.  To get the release name use::

  > helm ls

To easily upgrade the release use::

  > helm upgrade [RELEASE] [CHART]

To roll back to a previous release version use::

  > helm rollback [flags] [RELEASE] [REVISION]

For example, to upgrade the onap-so helm release to the latest SO container
release v1.1.2:

- Edit so values.yaml which is part of the chart
- Change "so: nexus3.onap.org:10001/openecomp/so:v1.1.1" to
  "so: nexus3.onap.org:10001/openecomp/so:v1.1.2"
- From the chart location run::

  > helm upgrade onap-so

The previous so pod will be terminated and a new so pod with an updated so
container will be created.

.. figure:: oomLogoV2-Delete.png
   :align: right

Delete
======

Existing deployments can be partially or fully removed once they are no longer
needed.  To minimize errors it is recommended that before deleting components
from a running deployment the operator perform a 'dry-run' to display exactly
what will happen with a given command prior to actually deleting anything.  For
example::

  > helm undeploy casablanca --dry-run

will display the outcome of deleting the 'casablanca' release from the
deployment.
To completely delete a release and remove it from the internal store enter::

  > helm undeploy casablanca --purge

One can also remove individual components from a deployment by changing the
ONAP configuration values.  For example, to remove `so` from a running
deployment enter::

  > helm undeploy casablanca-so --purge

will remove `so` as the configuration indicates it's no longer part of the
deployment. This might be useful if a one wanted to replace just `so` by
installing a custom version.