Skip to content

Commit

Permalink
WMS ID#11455 : Use Raft Replication with Distributed Database for Res…
Browse files Browse the repository at this point in the history 
…ilient Never-Down Apps (#626)

* WMS# 11455 RAFT Live Labs

Initial pull request for RAFT Replication Live Labs

* WMS ID#11455 RAFT Replication Update

WMS ID#11455 RAFT Replication Live labs Update

* WMS ID#11455 : RAFT Replication

Updates to one of the task in WMS ID#11455 : RAFT Replication

* WMS ID#11455 - Raft Replication Workshop

WMS ID#11455 - Raft Replication Workshop Updates

* WMS ID#11455 - Raft Replication-update

WMS ID#11455 - Raft Replication-update

* WMS ID #11455 - Learn how to use RAFT Replication to build resilient distributed databases (PR #341)

WMS ID #11455 - Learn how to use RAFT Replication to build resilient distributed databases (PR #341) -Updates requested from Zachary Talke <notifications@github.com>

* WMS ID#11455 : Learn how to use RAFT Replication to build resilient distributed -Update

WMS ID#11455 : Learn how to use RAFT Replication to build resilient distributed -Update (remove : *Free-tier* and *Paid Tenants* only)

* WMS ID #11455 - Learn how to use RAFT Replication to build resilient distributed databases (PR #341)

WMS ID #11455 - Learn how to use RAFT Replication to build resilient distributed databases (PR #341) - Update : removing (*Free-tier* and *Paid Tenants* only)

* WMS ID#11455 - Learn how to use RAFT Replication to build resilient distributed databases -Updates req. from Livelabs

WMS ID#11455 - Learn how to use RAFT Replication to build resilient distributed databases -Updates requested from Livelabs (Anoosha Pilli)

* WMS ID #11455 - Learn how to use RAFT Replication to build resilient distributed databases-Update PAR link

WMS ID #11455 - Learn how to use RAFT Replication to build resilient distributed databases-Update PAR link

* WMS ID #11455 Updates

Use-case updates to Raft Replication Live Labs

* Updates to WMS ID #11455

Updates to image names and sequence numbers in Raft Replication Live Labs

* Update: WMS ID #11455

Raft Replication live labs image names  correction

* Updates to WMS ID #11455

Raft Replication live labs - new images updated

* Updates to WMS ID #11455

Raft Replication live labs - topology lab update

* Updates to WMS ID #11455

WMS ID #11455 updates: Raft Replication Live Labs - New screenshots

* Updates to WMS ID #11455

WMS ID #11455 Raft Replication Live Labs
Updates to screenshots

* Updates to WMS ID #11455

Updates to WMS ID #11455 Raft Replication Live Labs - including cases, documentation links.

* Updates to WMS ID #11455

Updates to WMS ID #11455
par link updated and
manifest.json for sandbox has been updated

* WMS ID#11455 : Use Raft Replication with Distributed Database for Resilient Never-Down Apps

WMS ID#11455 : Updates to Raft Replication Live Labs - New images and adding Demo steps

* WMS ID#11455 : Use Raft Replication with Distributed Database for Resilient Never-Down Apps

WMS ID#11455 : Updates to  Raft Replication Live Labs - few steps updated

* WMS ID#11455 : Use Raft Replication with Distributed Database for Resilient Never-Down Apps

Updates to WMS ID#11455 : Use Raft Replication with Distributed Database for Resilient Never-Down Apps; Documentation updates

* Updates to WMS ID#11455 : Use Raft Replication with Distributed Database for Resilient Never-Down Apps

WMS ID#11455 : Use Raft Replication with Distributed Database for Resilient Never-Down Apps; updates to Appendix

* WMS ID#11455 : Use Raft Replication with Distributed Database for Resilient Never-Down Apps

Updates to WMS ID#11455
formatting required

* updates to WMS ID#11455

updates to WMS ID#11455

* updates to WMS ID#11455

updates to WMS ID#11455

* WMS ID#11455

Updates as suggested by Live Labs team -WMS ID#11455

---------

Co-authored-by: Anoosha Pilli <63126232+anooshapilli@users.noreply.github.com>
Co-authored-by: Dan Wiliams <127415766+dannymgithub@users.noreply.github.com>
  • Loading branch information
@deekshasehgal @anooshapilli @dannymgithub
3 people authored Aug 14, 2024
1 parent 0090da5 commit 22e5b68
Show file tree
Hide file tree
Showing 14 changed files with 396 additions and 438 deletions.
466 changes: 240 additions & 226 deletions sharding/raft23ai/advanced-use-cases/advanced-use-cases.md
View file

Large diffs are not rendered by default.

Binary file added sharding/raft23ai/advanced-use-cases/images/t3-2-auto-rebalance.png
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
4 changes: 0 additions & 4 deletions sharding/raft23ai/initialize-environment/initialize-environment.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,10 +6,6 @@ In this lab we will review and startup all components required to successfully r

*Estimated Lab Time:* 10 Minutes.

Watch the video for a quick walk through of the Initialize Environment lab.

[Initialize Environment lab](youtube:e3EXx3BMhec)

### Objectives
- Initialize the workshop environment.

Expand Down
Binary file removed sharding/raft23ai/topology/images/t3-6-exit-gsm1.png
View file
Binary file not shown.
Binary file added sharding/raft23ai/topology/images/terminal-1-oracle.png
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added sharding/raft23ai/topology/images/terminal-2-gsm.png
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added sharding/raft23ai/topology/images/terminal-3-appclient.png
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
205 changes: 134 additions & 71 deletions sharding/raft23ai/topology/topology.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -33,7 +33,33 @@ This lab assumes you have:

## Task 1: Check for containers in your VM

1. Open a terminal window and execute below as **opc** user.
1. Please open three terminal windows.

First terminal logged in as **oracle** user.

![<terminal_oracle>](./images/terminal-1-oracle.png " ")

Second Terminal switched to **GSM** level.

```
<copy>
sudo podman exec -i -t gsm1 /bin/bash
</copy>
```

![<terminal_gsm>](./images/terminal-2-gsm.png " ")

Third Terminal switched to **appclient** container.

```
<copy>
sudo podman exec -it appclient /bin/bash
</copy>
```

![<terminal_appclient>](./images/terminal-3-appclient.png " ")

2. Run the below command on terminal window that is logged in as as **oracle** user.

```
<copy>
Expand All @@ -53,7 +79,8 @@ Changes to data made by a DML are recorded in the Raft log. A commit record is a
For more details check [Raft Replication Configuration and Management] (https://docs.oracle.com/en/database/oracle/oracle-database/23/shard/raft-replication.html#GUID-AF14C34B-4F55-4528-8B28-5073A3BFD2BE)
1. Run in the terminal as **oracle** user and connect to the shard director server.
1. Run the below command to switch to **GSM**, if you are using a new terminal.
```
<copy>
sudo podman exec -i -t gsm1 /bin/bash
Expand All @@ -62,7 +89,7 @@ For more details check [Raft Replication Configuration and Management] (https://
![<connect to GSM1>](./images/t2-1-podman-gsm1.png " ")
2. Verify sharding topology using the **CONFIG** command.
2. Use the terminal window that is switched to **GSM**. Verify sharding topology using the **CONFIG** command.
```
<copy>
Expand Down Expand Up @@ -105,9 +132,11 @@ For more details check [Raft Replication Configuration and Management] (https://
## Task 3: Changing the Replication Unit Leader
Using SWITCHOVER RU, you can change which replica is the leader for the specified replication unit. The -db option makes the specified database the new leader of the given RU.
Using SWITCHOVER RU, you can change which replica is the leader for the specified replication unit.
The -shard option makes the replication unit member on the specified shard database the new leader of the given RU.
1. Run the below command on GSM1 to view the status of all the leaders
1. Run the below command on **GSM** terminal window to view the status of all the leaders
```
<copy>
Expand Down Expand Up @@ -154,29 +183,22 @@ Using SWITCHOVER RU, you can change which replica is the leader for the specifie
![<leader_status_after_change>](./images/t3-5-status-after-leader-change.png " ")
6. Exit from GSM1.
```
<copy>
exit
</copy>
```
![<t3-6-exit-gsm1>](./images/t3-6-exit-gsm1.png " ")
## Task 4: Run the workload
Please use the below steps to run the workload using the "app_schema" account with the available configuration files on the "appclient" container:
1. Switch to the "appclient" container
1. You can use the below command if you need to switch to appclient container in a new terminal window.
```
<copy>
sudo podman exec -it appclient /bin/bash
</copy>
```
![<appclient_container>](./images/t4-1-appclient-container.png " ")
2. Switch to the "oracle" user
2. Use the terminal window that is switched to the "appclient" container. Switch to the "oracle" user.
```
<copy>
Expand Down Expand Up @@ -208,8 +230,8 @@ Please use the below steps to run the workload using the "app_schema" account wi
![<run_workload>](./images/t4-4-run-workload.png " ")
5. During this time, you can continue to check the RU details from another session on the "gsm1" container from "gdsctl" prompt.
Notice that the log index is increasing as there are read and write operations are going on.
5. During this time, you can continue to check the RU details from another terminal window switched to **gsm** .
Notice that the log index is increasing as read and write operations are going on.
```
<copy>
Expand All @@ -227,7 +249,7 @@ Notice that the log index is increasing as there are read and write operations a
What happens when one of the available shard databases goes down or is taken down for maintenance?
Failover test by stopping shard1 to create shard1 down situation.
1. Run the below command as **oracle** user to check the status for all the containers.
1. You can run the below command in a terminal window logged in as **oracle** user to check the status for all the containers.
```
<copy>
Expand All @@ -249,14 +271,16 @@ Failover test by stopping shard1 to create shard1 down situation.
![<stop_shard_1>](./images/t5-2-stop-shard1.png " ")
3. Switch to GSM1 on another terminal session and check the status for RU's and you will see that database orcl1cdb_orcl1pdb is not present.
3. Below command can be used to switch to **GSM**, if you are using a new terminal.
```
<copy>
sudo podman exec -i -t gsm1 /bin/bash
</copy>
```
Run below in the terminal window that is switched to **GSM** and check the status of shards, RU's and you will see that database orcl1cdb_orcl1pdb is not present.
```
<copy>
gdsctl config shard
Expand All @@ -273,7 +297,8 @@ Failover test by stopping shard1 to create shard1 down situation.
You will see that shard1 down situation has no impact on the running workload.
4. Start the shard1 using the podman start command, to reflect that shard1 is joining back.
4. On a terminal window logged in as **oracle**.
Start the shard1 using the podman start command, to reflect that shard1 is joining back.
```
<copy>
Expand All @@ -284,14 +309,16 @@ You will see that shard1 down situation has no impact on the running workload.
![<start_shard1>](./images/t5-4-startup-shard1.png " ")
5. On a parallel session switch to GSM1, check the status of shard, RU's and see that shard1 has joined back.
5. You can use the below command as **oracle** to switch to **GSM**.
```
<copy>
sudo podman exec -i -t gsm1 /bin/bash
</copy>
```
On a terminal window switched to **GSM**, check the status of shard, RU's and see that shard1 has joined back.
```
<copy>
gdsctl config shard
Expand All @@ -308,6 +335,16 @@ You will see that shard1 down situation has no impact on the running workload.
You can stop the workload that ran in the previous task using Ctrl+C.
6. Run the below command in terminal that is switched to **GSM** to auto rebalance the leaders.
```
<copy>
gdsctl switchover ru -rebalance
</copy>
```
![<change_the_ru_leader>](./images/t3-2-auto-rebalance.png " ")
## Task 6: Raft Replication Demo with UI Application
Expand All @@ -329,125 +366,151 @@ Verify where the leadership is for User No. 1's shard in the Demo UI.
![<demo_ui_app_more_details>](./images/demo-ui-app-more-det-3.png " ")
4. Go to the terminal and stop that shard.
4. Go to the terminal window logged in as **oracle** and stop that shard.
```
<copy>
sudo podman stop shard3
</copy>
```
You can use below command to switch to **GSM**.
```
<copy>
sudo podman exec -i -t gsm1 /bin/bash
</copy>
```
Run the below command in terminal window that is switched to **GSM**, to see the status of the shard.
```
<copy>
gdsctl config shard
</copy>
```
![<demo_ui_app_stop_shard>](./images/demo-ui-app-stop-shard-4.png " ")
![<demo_ui_app_stop_shard>](./images/demo-ui-app-stop-shard-4.png " ")
5. Return to the Demo UI App browser window on "More Details" tab and click on "Refresh" button to observe that the leadership has automatically moved to another shard, indicating re-routing of the request.
![<demo_ui_app_refresh_more_details>](./images/demo-ui-app-refresh-more-det-5.png " ")
![<demo_ui_app_refresh_more_details>](./images/demo-ui-app-refresh-more-det-5.png " ")
6. Go to the first update tab in Demo UI Application and change the class.
![<demo_ui_app_update_class>](./images/demo-ui-app-update-class-6.png " ")
![<demo_ui_app_update_class>](./images/demo-ui-app-update-class-6.png " ")
Go to the next update tab and refresh it to see that change in class.
Click on the next update tab and refresh it to see that change in class.
![<demo_ui_app_class_updated>](./images/demo-ui-app-class-updated-6.png " ")
![<demo_ui_app_class_updated>](./images/demo-ui-app-class-updated-6.png " ")
7. Run the Workload and check the count of customers in Demo UI App.
Check the count in browser window before running the workload.
Check the count in browser window before running the workload.
![<demo_ui_check_count_before_workload>](./images/demo-ui-app-checkcount-before-workload-7.png " ")
![<demo_ui_check_count_before_workload>](./images/demo-ui-app-checkcount-before-workload-7.png " ")
Note that while the shard is stopped, you can still run the workload.
Note that while the shard is stopped, you can still run the workload.
Switch to the "appclient" container
If you are using a new terminal, you can use below command as **oracle** to switch to **appclient** container and the switch to the "oracle" user and then change the path to $DEMO_MASTER location.
```
<copy>
sudo podman exec -it appclient /bin/bash
</copy>
```
```
<copy>
sudo podman exec -it appclient /bin/bash
</copy>
```
Switch to the "oracle" user
```
<copy>
su - oracle
</copy>
```
```
<copy>
su - oracle
</copy>
```
```
<copy>
cd $DEMO_MASTER
pwd
ls -rlt
</copy>
```
Change the path to $DEMO_MASTER location
Run the workload using the below command in a terminal window that is switched to **appclient** container.
```
<copy>
cd $DEMO_MASTER
pwd
ls -rlt
</copy>
```
```
<copy>
sh run.sh demo
</copy>
```
Run the workload using the below command
![<demo_ui_app_run_workload_a>](./images/demo-ui-app-run-workload-7a.png " ")
```
<copy>
sh run.sh demo
</copy>
```
![<demo_ui_app_run_workload_a>](./images/demo-ui-app-run-workload-7a.png " ")
![<demo_ui_app_run_workload>](./images/demo-ui-app-run-workload-7b.png " ")
![<demo_ui_app_run_workload>](./images/demo-ui-app-run-workload-7b.png " ")
Refresh the browser window for Demo UI application and you will observe that the count is increasing in the Demo UI even though the shard is stopped.
Refresh the browser window for Demo UI application and you will observe that the count is increasing in the Demo UI even though the shard is stopped.
![<demo_ui_app_inc_after_workload>](./images/demo-ui-app-count-inc-after-workload-7c.png " ")
![<demo_ui_app_inc_after_workload>](./images/demo-ui-app-count-inc-after-workload-7c.png " ")
8. Stop the Workload.
You can stop the workload that ran in the previous task using Ctrl+C.
8. Stop the Workload.
9. Run the below command in terminal that is switched to **GSM** to auto rebalance the leaders.
You can stop the workload that ran in the previous task using Ctrl+C.
```
<copy>
gdsctl switchover ru -rebalance
</copy>
```
![<change_the_ru_leader>](./images/t3-2-auto-rebalance.png " ")
9. Now, we are going to perform delete operation through Demo UI Application.
10. Now, we are going to perform delete operation through Demo UI Application.
Take a look at the count before clicking on delete button.
Take a look at the count before clicking on delete button.
![<demo_ui_app_count_before_del>](./images/demo-ui-app-count-before-delete-9.png " ")
![<demo_ui_app_count_before_del>](./images/demo-ui-app-count-before-delete-9.png " ")
Click on Delete button in the browser window and delete few users. You will notice that the count is decreasing.
s
![<demo_ui_app_after_del_count>](./images/demo-ui-app-after-del-count-9a.png " ")
Click on Delete button in the browser window and delete few users. You will notice that the count is decreasing.
![<demo_ui_app_after_del_count>](./images/demo-ui-app-after-del-count-9a.png " ")
11. Start the shard and see that the shard1 is joining back.
11. Run the below command in terminal window logged in as **oracle**. Start the shard and see that the shard1 is joining back.
```
<copy>
sudo podman start shard3
</copy>
```
![<demo_ui_app_start_shard>](./images/demo-ui-app-start-shard-11.png " ")
Run the below command in terminal that is switched to **GSM** to check the status of the shard.
```
<copy>
gdsctl config shard
</copy>
```
![<demo_ui_app_start_shard>](./images/demo-ui-app-start-shard-11.png " ")
12. Run the below command in terminal that is switched to **GSM** to auto rebalance the leaders.
```
<copy>
gdsctl switchover ru -rebalance
</copy>
```
![<change_the_ru_leader>](./images/t3-2-auto-rebalance.png " ")
You may now proceed to the next lab.
## Acknowledgements
* **Authors** - Deeksha Sehgal, Oracle Globally Distributed Database Database, Product Management, Senior Product Manager
* **Contributors** - Pankaj Chandiramani, Shefali Bhargava, Ajay Joshi, Jyoti Verma
* **Last Updated By/Date** - Ajay Joshi, Oracle Globally Distributed Database, Product Management, Consulting Member of Technical Staff, July 2024
* **Last Updated By/Date** - Deeksha Sehgal, Oracle Globally Distributed Database, Product Management, Senior Product Manager, Aug 2024
Loading

0 comments on commit 22e5b68

Please sign in to comment.