Does JSM Datacenter supports database partitioning?

[Jose Hernandez]

Hi,

Does JSM Datacenter supports database partitioning for performance enhancement?

Thanks in advance,

Regards

[Sagar Mahajan · Accepted solution]

Hi @Jose Hernandez database partitioning you mean reader and writer node for JSM to offload the read and write traffic towards the database from application?

If so, Then Jira supports the database with one writer and multiple reader nodes but you can only configure the writer node in configuration there is no such configuration option for readers.

You can check the below document for more details about configuring Jira with AWS Aurora database.

jira-software-data-center-on-the-aws-cloud 

Let me know if you looking for something else.

[Jose Hernandez]

Hi Sagar,

Thanks for your answer! Yes, I meant to spread each database partition over multiple nodes. I understand that this is supported, as long as there is just one writer node. With "you can only configure the writer node in configuration" you mean it is only possible to configure the writer node using the Jira Web UI?

[Sagar Mahajan]

Yes, we can only configure writer endpoint from Jira UI.

I am not sure but looks like AWS internally redirect the read requests to the reader node as I can see some CPU usage and load on the reader node in my setup.

Unlike Bugzilla where we can configure reader and writer node separately we don't have such options in Jira.

You can select the appropriate database class if your planning to use AWS database that supports good amount of database connection pools so that multiple requests will be served per second by the database.

I hope this helps you define your JSM architecture.