Sync docs from Discourse (#589)

Co-authored-by: GitHub Actions <41898282+github-actions[bot]@users.noreply.github.com>
canonical · Aug 20, 2024 · 9c0fbae · 9c0fbae
1 parent 3316ac8
commit 9c0fbae
Show file tree

Hide file tree

Showing 40 changed files with 399 additions and 376 deletions.
diff --git a/docs/explanation/e-legacy-charm.md b/docs/explanation/e-legacy-charm.md
@@ -1,4 +1,4 @@
-## Charm types "legacy" vs "modern"
+## Charm types: "legacy" vs "modern"
 There are [two types of charms](https://juju.is/docs/sdk/charm-taxonomy#heading--charm-types-by-generation) stored under the same charm name `postgresql`:
 
 1. [Reactive](https://juju.is/docs/sdk/charm-taxonomy#heading--reactive)  charm in the channel `latest/stable` (called `legacy`)

diff --git a/docs/explanation/e-statuses.md b/docs/explanation/e-statuses.md
@@ -9,6 +9,7 @@ The charm follows [standard Juju applications statuses](https://juju.is/docs/olm
 | **active** | any | Normal charm operations | No actions required |
 | **waiting** | any | Charm is waiting for relations to be finished | No actions required |
 | **maintenance** | any | Charm is performing the internal maintenance (e.g. cluster re-configuration, upgrade, ...) | No actions required |
+| **blocked** | the S3 repository has backups from another cluster | The bucket contains foreign backup. To avoid accident DB corruption, use clean bucket. The cluster identified by Juju app name + DB UUID. | Chose/change the new S3 [bucket](https://charmhub.io/s3-integrator/configuration#bucket)/[path](https://charmhub.io/s3-integrator/configuration#path) OR clean the current one. |
 | **blocked** | failed to update cluster members on member | TODO: error/retry? | |
 | **blocked** | failed to install snap packages | There are issues with the network connection and/or the Snap Store | Check your internet connection and https://status.snapcraft.io/. Remove the application and when everything is ok, deploy the charm again |
 | **blocked** | failed to patch snap seccomp profile | The charm failed to patch one issue that happens when pgBackRest restores a backup (this blocked status should be removed when https://github.com/pgbackrest/pgbackrest/releases/tag/release%2F2.46 is added to the snap) | Remove the unit and add it back again |

diff --git a/docs/how-to/h-async/h-async-integrate.md → docs/how-to/h-async-integrate.md b/docs/how-to/h-async/h-async-integrate.md → docs/how-to/h-async-integrate.md
@@ -1,9 +1,4 @@
 # Integrate with a client application
-[note type="caution"]
-This is an internal article. **Do not use it in production!** 
-
-Contact the [Canonical Data Platform team](https://chat.charmhub.io/charmhub/channels/data-platform) if you are interested in this topic.
-[/note]
 
 This guide will show you how to integrate a client application with a cross-regional async setup using an example PostgreSQL deployment with two servers: one in Rome and one in Lisbon.
 
@@ -13,13 +8,13 @@ This guide will show you how to integrate a client application with a cross-regi
   * Refer to the page [How to set up clusters](/t/13991)
 
 ## Summary
-* [Configure database endpoints](#heading--configure-endpoints)
-* [Internal client](#heading--internal-client)
-* [External client](#heading--external-client)
+* [Configure database endpoints](#configure-database-endpoints)
+* [Internal client](#internal-client)
+* [External client](#external-client)
 
 ---
 
-<a href="#heading--configure-endpoints"><h2 id="heading--configure-endpoints"> Configure database endpoints </h2></a>
+## Configure database endpoints
 
 To make your database available to a client application, you must first offer and consume database endpoints.
 
@@ -46,7 +41,7 @@ juju consume rome.db1database
 juju consume lisbon.db2database
 ```
 
-<a href="#heading--internal-client"><h2 id="heading--internal-client"> Internal client </h2></a>
+## Internal client
 
 If the client application is another charm, deploy them and connect them with `juju integrate`.
 
@@ -62,7 +57,7 @@ juju integrate postgresql-test-app:first-database pgbouncer
 juju integrate pgbouncer db1database
 ```
 
-<a href="#heading--external-client"><h2 id="heading--external-client"> External client </h2></a>
+## External client
 
 If the client application is external, they must be integrated via the [`data-integrator` charm](https://charmhub.io/data-integrator).
 

diff --git a/.../how-to/h-async/h-async-remove-recover.md → docs/how-to/h-async-remove-recover.md b/.../how-to/h-async/h-async-remove-recover.md → docs/how-to/h-async-remove-recover.md
@@ -1,9 +1,4 @@
 # Remove or recover a cluster
-[note type="caution"]
-This is an internal article. **Do not use it in production!** 
-
-Contact the [Canonical Data Platform team](https://chat.charmhub.io/charmhub/channels/data-platform) if you are interested in this topic.
-[/note]
 
 This guide will cover how to manage clusters using an example PostgreSQL deployment with two servers: one in Rome and one in Lisbon.
 
@@ -13,16 +8,16 @@ This guide will cover how to manage clusters using an example PostgreSQL deploym
   * Refer to the page [How to set up clusters](/t/13991)
 
 ## Summary
-* [Switchover](#heading--switchover)
-* [Detach a cluster](#heading--detach)
-  * [Reuse a detached cluster](#heading--reuse)
-  * [Remove a detached cluster](#heading--remove)
-* [Recover a cluster](#heading--recover)
+* [Switchover](#switchover)
+* [Detach a cluster](#detach-a-cluster)
+  * [Reuse a detached cluster](#reuse-a-detached-cluster)
+  * [Remove a detached cluster](#remove-a-detached-cluster)
+* [Recover a cluster](#recover-a-cluster)
 
 <!-- TODO: Rethink sections, especially "recover" -->
 ---
 
-<a href="#heading--switchover"><h2 id="heading--switchover"> Switchover </h2></a>
+## Switchover
 
 If the primary cluster fails or is removed, it is necessary to appoint a new cluster as primary.
 
@@ -32,7 +27,7 @@ To switchover and promote `lisbon` to primary, one would run the command:
 juju run -m lisbon db2/leader promote-to-primary
 ```
 
-<a href="#heading--detach"><h2 id="heading--detach"> Detach a cluster </h2></a>
+## Detach a cluster
 
 Clusters in an async replica set can be detached. The detached cluster can then be either removed or reused.
 
@@ -44,22 +39,22 @@ juju remove-relation -m lisbon replication-offer db2:replication
 
 The command above will move the `rome` cluster into a detached state (`blocked`) keeping all the data in place.
 
-<a href="#heading--reuse"><h3 id="heading--reuse"> Reuse a detached cluster </h3></a>
+### Reuse a detached cluster
 
 The following command creates a new cluster in the replica set from the detached `rome` cluster, keeping its existing data in use:
 
 ```shell
 juju run -m rome db1/leader promote-to-primary
 ```
-<a href="#heading--remove"><h3 id="heading--remove"> Remove a detached cluster </h3></a>
+### Remove a detached cluster
 
 The following command removes the detached `rome` cluster and **destroys its stored data** with the optional `--destroy-storage` flag:
 
 ```shell
 juju remove-application -m rome db1 --destroy-storage
 ```
 
-<a href="#heading--recover"><h2 id="heading--recover"> Recover a cluster </h2></a>
+## Recover a cluster
 
 **If the integration between clusters was removed** and one side went into a  `blocked` state, integrate both clusters again and call the `promote-cluster` action to restore async replication - similar to the "Reuse a detached cluster" step above.
 

diff --git a/docs/how-to/h-async-set-up.md b/docs/how-to/h-async-set-up.md
@@ -0,0 +1,75 @@
+# Set up clusters for cross-regional async replication
+
+Cross-regional (or multi-server) asynchronous replication focuses on disaster recovery by distributing data across different servers. 
+
+This guide will show you the basics of initiating a cross-regional async setup using an example PostgreSQL deployment with two servers: one in Rome and one in Lisbon.
+
+## Summary
+* [Deploy](#deploy)
+* [Offer](#offer)
+* [Consume](#consume)
+* [Promote or switchover a cluster](#promote-or-switchover-a-cluster)
+* [Scale a cluster](#scale-a-cluster)
+
+---
+
+## Deploy
+
+To deploy two clusters in different servers, create two juju models - one for the `rome` cluster, one for the `lisbon` cluster. In the example below, we use the config flag `profile=testing` to limit memory usage.
+
+```shell
+juju add-model rome 
+juju add-model lisbon
+
+juju switch rome # active model must correspond to cluster
+juju deploy postgresql db1 --channel=14/edge/async-replication --config profile=testing --base [email protected]
+
+juju switch lisbon 
+juju deploy postgresql db2 --channel=14/edge/async-replication --config profile=testing --base [email protected]
+```
+
+## Offer
+
+[Offer](https://juju.is/docs/juju/offer) asynchronous replication in one of the clusters.
+
+```shell
+juju switch rome
+juju offer db1:async-primary async-primary
+``` 
+
+## Consume
+
+Consume asynchronous replication on planned `Standby` cluster (Lisbon):
+```shell
+juju switch lisbon
+juju consume rome.async-primary
+juju integrate async-primary db2:async-replica
+``` 
+
+## Promote or switchover a cluster
+
+To define the primary cluster, use the `promote-cluster` action.
+
+```shell
+juju run -m rome db1/leader promote-cluster
+```
+
+To switchover and use `lisbon` as the primary instead, run
+
+```shell
+juju run -m lisbon db2/leader promote-cluster force-promotion=true
+```
+
+## Scale a cluster
+
+The two clusters work independently, which means that it’s possible to scale each cluster separately. The `-m` flag defines the target of this action, so it can be performed within any active model. 
+
+For example:
+
+```shell
+juju add-unit db1 -n 2 -m rome
+juju add-unit db2 -n 2 -m lisbon
+```
+[note]
+**Note:** Scaling is possible before and after the asynchronous replication is established/created.
+[/note]
diff --git a/docs/how-to/h-async/h-async-set-up.md b/docs/how-to/h-async/h-async-set-up.md
diff --git a/docs/how-to/h-backups/h-configure-s3-aws.md → docs/how-to/h-configure-s3-aws.md b/docs/how-to/h-backups/h-configure-s3-aws.md → docs/how-to/h-configure-s3-aws.md
diff --git a/...ow-to/h-backups/h-configure-s3-radosgw.md → docs/how-to/h-configure-s3-radosgw.md b/...ow-to/h-backups/h-configure-s3-radosgw.md → docs/how-to/h-configure-s3-radosgw.md
diff --git a/docs/how-to/h-backups/h-create-backup.md → docs/how-to/h-create-backup.md b/docs/how-to/h-backups/h-create-backup.md → docs/how-to/h-create-backup.md
diff --git a/docs/how-to/h-set-up/h-deploy-lxd.md → docs/how-to/h-deploy-lxd.md b/docs/how-to/h-set-up/h-deploy-lxd.md → docs/how-to/h-deploy-lxd.md
diff --git a/docs/how-to/h-set-up/h-deploy-maas.md → docs/how-to/h-deploy-maas.md b/docs/how-to/h-set-up/h-deploy-maas.md → docs/how-to/h-deploy-maas.md