Skip to content

Latest commit

 

History

History
559 lines (377 loc) · 22.5 KB

README.rst

File metadata and controls

559 lines (377 loc) · 22.5 KB
Raw

PAVICS Components

Contents

Scheduler

This component provides automated unattended continuous deployment for the "PAVICS stack" (all the git repos in var AUTODEPLOY_EXTRA_REPOS), for the tutorial notebooks on the Jupyter environment and for the automated deployment itself.

It can also be used to schedule other tasks on the PAVICS physical host.

Everything is dockerized, the deployment runs inside a container that will update all other containers.

Automated unattended continuous deployment means if code change in the remote repo, matching the same currently checkout branch (ex: config changes, docker-compose.yml changes) a deployment will be performed automatically without human intervention.

The trigger for the deployment is new code change on the server on the current branch (PR merged, push). New code change locally will not trigger deployment so local development workflow is also supported.

Multiple remote repos are supported so the "PAVICS stack" can be made of multiple checkouts for modularity and extensibility. The autodeploy will trigger if any of the checkouts (configured in AUTODEPLOY_EXTRA_REPOS) is not up-to-date with its remote repo.

A suggested "PAVICS stack" is made of at least 2 repos, this repo and another private repo containing the source controlled env.local file and any other docker-compose override for true infrastructure-as-code.

Note: there are still cases where a human intervention is needed. See note in script deploy.sh (:download:`download <../deployment/deploy.sh>`).

Usage

Given the unattended nature, there is no UI. Logs are used to keep trace.

  • /var/log/PAVICS/autodeploy.log is for the PAVICS deployment.
  • /var/log/PAVICS/notebookdeploy.log is for the tutorial notebooks deployment.
  • logrotate is enabled for /var/log/PAVICS/*.log to avoid filling up the disk. Any new .log files in that folder will get logrotate for free.

How to Enable the Component

Old way to deploy the automatic deployment

Superseded by this new scheduler component. Keeping for reference only.

Doing it this old way do not need the scheduler component but lose the ability for the autodeploy system to update itself.

Configure logrotate for all following automations to prevent disk full:

deployment/install-logrotate-config .. $USER

To enable continuous deployment of PAVICS:

deployment/install-automated-deployment.sh .. $USER [daily|5-mins]
# read the script for more options/details

If you want to manually force a deployment of PAVICS (note this might not use latest version of deploy.sh script (:download:`download <../deployment/deploy.sh>`):

deployment/deploy.sh .
# read the script for more options/details

To enable continuous deployment of tutorial Jupyter notebooks:

deployment/install-deploy-notebook .. $USER
# read the script for more details

To trigger tutorial Jupyter notebooks deploy manually:

# configure logrotate before because this script will log to
# /var/log/PAVICS/notebookdeploy.log

deployment/trigger-deploy-notebook
# read the script for more details

Migrating to the new mechanism requires manual deletion of all the artifacts created by the old install scripts: sudo rm /etc/cron.d/PAVICS-deploy /etc/cron.hourly/PAVICS-deploy-notebooks /etc/logrotate.d/PAVICS-deploy /usr/local/sbin/triggerdeploy.sh. Both can not co-exist at the same time.

Comparison between the old and new autodeploy mechanism

Maximum backward-compatibility has been kept with the old install scripts style:

  • Still log to the same existing log files under /var/log/PAVICS.
  • Old single ssh deploy key is still compatible, but the new mechanism allows for different ssh deploy keys for each extra repos (again, public repos should use https clone path to avoid dealing with ssh deploy keys in the first place).
  • Old install scripts are kept and can still deploy the old way.

Features missing in old install scripts or how the new mechanism improves on the old install scripts:

  • Autodeploy of the autodeploy itself ! This is the biggest win. Previously, if triggerdeploy.sh (:download:`download <../deployment/triggerdeploy.sh>`) or the deployed /etc/cron.hourly/PAVICS-deploy-notebooks script changes, they have to be deployed manually. It's very annoying. Now they are volume-mount in so are fresh on each run.
  • env.local now drives absolutely everything, source control that file and we've got a true DevOPS pipeline.
  • Configurable platform and notebook autodeploy frequency. Previously, this means manually editing the generated cron file, less ideal.
  • Do not need any support on the local host other than docker and docker-compose. cron/logrotate/git/ssh versions are all locked-down in the docker images used by the autodeploy. Recall previously we had to deal with git version too old on some hosts.
  • Each cron job run in its own docker image meaning the runtime environment is traceable and reproducible.
  • The newly introduced scheduler component is made extensible so other jobs can added into it as well (ex: backup), via env.local, which should be source controlled, meaning all surrounding maintenance related tasks can also be traceable and reproducible.

How to test platform autodeploy is not broken by a PR

There are 2 tests that need to be performed:

  • Can autodeploy deploy the PR from master branch, the stable reference point?
    • This could fail if some changes in the PR are incompatible with autodeploy. For example: ./pavics-compose.sh calls some binaries that do not exist in the autodeploy docker image.
  • Can autodeploy be triggered again successfully, after the PR is live?
    • This could fail if the PR renamed some files and forgot to add the old file names to a .gitignore file. Then old file names will appear as new uncommitted files and autodeploy will halt because it expects a clean working directory.

Here is a sample setup to test autodeploy:

  • Have 2 checkout directories. One is for starting the stack using ./pavics-compose.sh, the other one is to push new bogus changes to trigger the autodeploy mechanism.
# this one for running pavics-compose.sh
git clone [email protected]:bird-house/birdhouse-deploy.git birdhouse-deploy

# this one for triggering autodeploy
git clone [email protected]:bird-house/birdhouse-deploy.git birdhouse-deploy-trigger
  • Set AUTODEPLOY_PLATFORM_FREQUENCY in env.local to a very frequent value so you do not have to wait too long for autodeploy to trigger.
# go to the main checkout
cd birdhouse-deploy/birdhouse

# ensure the scheduler component is enabled, otherwise autodeploy will not work
echo 'export EXTRA_CONF_DIRS="$EXTRA_CONF_DIRS ./components/scheduler" >> env.local

# set AUTODEPLOY_PLATFORM_FREQUENCY
# can set to more frequent than 5 minutes if your machine is capable enough
echo 'export AUTODEPLOY_PLATFORM_FREQUENCY="@every 5m"' >> env.local

# if scheduler container already running:
# recreate scheduler container for new AUTODEPLOY_PLATFORM_FREQUENCY to be effective
./pavics-compose.sh stop scheduler && ./pavics-compose.sh rm -vf scheduler && ./pavics-compose.sh up -d

# if scheduler container not running yet: start the newly added scheduler component
./pavics-compose.sh up -d
  • Create a ${USER}-test branch so you can add bogus commits without affecting your real PR. Set up your main checkout (birdhouse-deploy) to track that test branch so it will detect new changes on the test branch and trigger the autodeploy.
# go to the main checkout
cd birdhouse-deploy/birdhouse

# initially create the ${USER}-test branch from master
# the ${USER} prefix is to avoid name clash if another user is also testing autodeploy
git checkout master
git pull
git checkout -b ${USER}-test
git push -u ${USER}-test

# ensure your runnings code is at "master" and is working correctly
# if you do not have a working baseline, you will not know if the breakage is due to autodeploy or your code
./pavics-compose.sh up -d
  • Test scenario 1, from master to your PR
# go to the other checkout to trigger autodeploy
cd birdhouse-deploy-trigger/birdhouse

# set branch ${USER}-test to the same commit as your PR, this will trigger autodeploy from master to your PR
git pull
git checkout ${USER}-test
git reset --hard YOUR_PR_BRANCH
git push

# now that the remote "${USER}-test" branch differs from the local "${USER}-test" branch in the birdhouse-deploy repo,
# the autodeploy mechanism will detect that the remote branch has changed and attempt to update the local branch

# follow logs, check for errors
tail -f /var/log/PAVICS/autodeploy.log

# each autodeploy trigger will start the log with
#   ==========
#   triggerdeploy START_TIME=2023-06-15T05:07:01+0000

# each autodeploy trigger will end the log with
#   triggerdeploy finished START_TIME=2023-06-15T05:07:01+0000
#   triggerdeploy finished   END_TIME=2023-06-15T05:07:06+0000

# do spot checks in the log, run Jenkins on your deployment if needed
  • Test scenario 2, from your PR to later changes
# go to the other checkout to trigger autodeploy
cd birdhouse-deploy-trigger/birdhouse

# add any bogus commit to trigger autodeploy again
echo >> README.rst
git add README.rst
git commit -m "trigger autodeploy"
git push

# now that the remote "${USER}-test" branch differs from the local "${USER}-test" branch in the birdhouse-deploy repo,
# the autodeploy mechanism will detect that the remote branch has changed and attempt to update the local branch

# follow logs, check for errors
tail -f /var/log/PAVICS/autodeploy.log
  • Test done, clean up the bogus ${USER}-test branch and optionally relax AUTODEPLOY_PLATFORM_FREQUENCY
# go to the other checkout to trigger autodeploy
cd birdhouse-deploy-trigger/birdhouse

# go to master so we can delete the ${USER}-test branch
git checkout master
git push origin --delete ${USER}-test
git branch -D ${USER}-test

# go to the main checkout
cd birdhouse-deploy/birdhouse

# go to YOUR_PR_BRANCH so we can delete the ${USER}-test branch
git checkout YOUR_PR_BRANCH
git branch -D ${USER}-test

# edit env.local and change AUTODEPLOY_PLATFORM_FREQUENCY to something less frequent to save your cpu
# do not remove the scheduler component from the stack yet or the next command will fail

# recreate scheduler container for new AUTODEPLOY_PLATFORM_FREQUENCY to be effective
./pavics-compose.sh stop scheduler && ./pavics-compose.sh rm -vf scheduler && ./pavics-compose.sh up -d

# optionally edit env.local to remove the scheduler component from the stack
# then remove the running scheduler container
./pavics-compose.sh up -d --remove-orphans

Monitoring

This component provides monitoring and alerting for the PAVICS physical host and containers.

Prometheus stack is used:

  • Node-exporter to collect host metrics.
  • cAdvisor to collect containers metrics.
  • Prometheus to scrape metrics, to store them and to query them.
  • AlertManager to manage alerts: deduplicate, group, route, silence, inhibit.
  • Grafana to provide visualization dashboard for the metrics.

Usage

The paths above are by default only accessible to a user logged in to magpie as an administrator or as a member of group monitoring. These routes provide sensitive information about the birdhouse-deploy software stack and the machine that it is running on. It is highly discouraged to make these routes available to anyone who does not have proper access permissions.

Add existing users to the monitoring group to allow them access to the various monitoring WebUI. This way, we do not need to share the admin user account and do not have to add them to the administrators group, which would give them too much permissions.

How to Enable the Component

  • Edit env.local (a copy of env.local.example (:download:`download <../env.local.example>`))
    • Add "./components/monitoring" to EXTRA_CONF_DIRS
    • Set GRAFANA_ADMIN_PASSWORD to login to Grafana
    • Set ALERTMANAGER_ADMIN_EMAIL_RECEIVER for receiving alerts
    • Set SMTP_SERVER for sending alerts
    • Optionally set
      • ALERTMANAGER_EXTRA_GLOBAL to further configure AlertManager
      • ALERTMANAGER_EXTRA_ROUTES to add more routes than email notification
      • ALERTMANAGER_EXTRA_INHIBITION to disable rule from firing
      • ALERTMANAGER_EXTRA_RECEIVERS to add more receivers than the admin emails
    • Alert thresholds can be customized by setting the various PROMETHEUS_*_ALERT vars in env.local. The list of PROMETHEUS_*_ALERT vars are in monitoring_default.env (:download:`download <monitoring/default.env>`).

Grafana Dashboard

monitoring/images/grafana-dashboard.png

For host, using Node-exporter to collect metrics:

  • uptime
  • number of container
  • used disk space
  • used memory, available memory, used swap memory
  • load
  • cpu usage
  • in and out network traffic
  • disk I/O

For each container, using cAdvisor to collect metrics:

  • in and out network traffic
  • cpu usage
  • memory and swap memory usage
  • disk usage

Useful visualisation features:

  • zoom in one graph and all other graph update to match the same "time range" so we can correlate event
  • view each graph independently for more details
  • mouse over each data point will show value at that moment

Prometheus Alert Rules

monitoring/images/prometheus-alert-rules.png

AlertManager for Alert Dashboard and Silencing

monitoring/images/alertmanager-dashboard.png

monitoring/images/alertmanager-silence-alert.png

Customizing the Component

  • To add more Grafana dashboard, volume-mount more *.json files to the grafana container.
  • To add more Prometheus alert rules, volume-mount more *.rules files to the prometheus container.
  • To disable existing Prometheus alert rules, add more Alertmanager inhibition rules using ALERTMANAGER_EXTRA_INHIBITION via env.local file.
  • Other possible Alertmanager configs via env.local: ALERTMANAGER_EXTRA_GLOBAL, ALERTMANAGER_EXTRA_ROUTES (can route to Slack or other services accepting webhooks), ALERTMANAGER_EXTRA_RECEIVERS.

Weaver

By enabling this component, the Weaver service will be integrated into the stack.

This component offers OGC API - Processes interface to WPS components (a.k.a WPS-REST bindings and WPS-T (Transactional) support). This provides a RESTful JSON interface with asynchronous WPS processes execution over remote instances. Other WPS components of the birdhouse stack (finch, flyingpigeon, etc.) will also all be registered under Weaver in order to provide a common endpoint to retrieve all available processes, and dispatch their execution to the corresponding service. Finally, Weaver also adds Docker image execution capabilities as a WPS process, allowing deployment and execution of custom applications and workflows.

weaver/images/component-diagram.png

Usage

Once this component is enabled, Weaver will be accessible at https://<PAVICS_FQDN_PUBLIC>/weaver endpoint, where PAVICS_FQDN_PUBLIC is defined in your env.local file.

Full process listing (across WPS providers) should be available using request:

GET https://<PAVICS_FQDN_PUBLIC>/weaver/processes?providers=true

Please refer to the Weaver OpenAPI for complete description of available requests. This description will also be accessible via https://<PAVICS_FQDN_PUBLIC>/weaver/api once the instance is started.

For any specific details about Weaver configuration parameters, functionalities or questions, please refer to its documentation.

How to Enable the Component

  • Edit env.local (a copy of env.local.example)
    • Add ./components/weaver to EXTRA_CONF_DIRS.
    • Component birdhouse/optional-components/all-public-access should also be enabled to ensure that Weaver can request GetCapabilities of every WPS provider to be registered. Publicly inaccessible services will not succeed registration and will not provide the WPS-REST interface.

Customizing the Component

  • Edit env.local (a copy of env.local.example)
    • Optionally, set any additional environment variable overrides amongst values defined in weaver/default.env.
    • Optionally, mount any additional Weaver-specific configuration files (see contents of birdhouse/components/weaver/config/weaver) if extended functionalities need to be defined. Further docker-compose-extra.yml could be needed to define any other volumes entries where these component would need to be mounted to.

Cowbird

Cowbird is a middleware that manages interactions between various birds of the bird-house stack.

It relies on the existence of other services under a common architecture, but applies changes to the resources under those services such that the complete ecosystem can seamlessly operate together (see Components Diagram).

The code of this service is located in Ouranosinc/cowbird. Its documentation is provided on ReadTheDocs.

Operations Performed by Cowbird

  • Synchronize Magpie user and group permissions between "corresponding files" located under different services. For example, THREDDS user-workspace files visualized in the catalog will be accessible by the same user under the corresponding user-workspace under GeoServer.
  • Synchronize Weaver endpoints to retrieve equivalent definitions under various paths and access to generated WPS outputs following a job execution by a given user.
  • Synchronize permissions between API endpoints and local storage files.
  • Synchronize permissions and references based on event triggers and request callbacks.

Usage

Cowbird is intended to work on its own, behind the scene, to apply any required resource synchronization between the various services of the platform when changes are detected. Therefore, it does not require any explicit interaction from users.

In case the platform maintainer desires to perform manual syncing operations with Cowbird, its REST API should be used. It will be accessible under https://{PAVICS_FQDN_PUBLIC}/cowbird and details of available endpoints will be served under /cowbird/api. Note that Magpie administrator credentials will be required to access those endpoints.

How to Enable the Component

  • Edit env.local (a copy of env.local.example)
  • Add ./components/cowbird to EXTRA_CONF_DIRS.

Customizing the Component

Cowbird can be affected by multiple variables defined globally on the stack (i.e.: env.local, a copy of env.local.example). It also considers variables of other services such as THREDDS, GeoServer, Magpie, etc. in order to perform required interactions between them.

By default, variables defined in cowbird/default.env will be used unless overridden in env.local. To apply changes define your custom values in env.local directly.

STAC

STAC is the common name of the REST API that implements the STAC specification, common representation of geospatial information.

Usage

The STAC API can be browsed via the stac-browser component. By default, the browser will point to the STAC API exposed by the current stack instance. Once this component is enabled, STAC API will be accessible at https://<PAVICS_FQDN_PUBLIC>/stac endpoint and the STAC browser will be available at https://<PAVICS_FQDN_PUBLIC>/stac-browser endpoint. In order to make the STAC browser the default entrypoint, define the following in the env.local file:

export PROXY_ROOT_LOCATION="return 302 https://\$host/stac-browser;"

Here is a sample search query using a CLI:

.. code-block:: shell
pip install pystac-client stac-client search $PAVIS_FQDN/stac -q "variable_id=txgt_32" "scenario=ssp585"

Calls to the STAC API pass through Twitcher in order to validate authorization. Unauthenticated users will have read-only access by default to STAC API resources while members of the stac-admin group can create and modify resources. STAC Browser is not protected by any authorization mechanism.

How to Enable the Component

  • Edit env.local (a copy of env.local.example)
  • Add ./optional-components/stac to EXTRA_CONF_DIRS.