Arima was recently hired as a Jira administrator for a large company who hasn’t had any previous dedicated Jira administration staff.
The instance he’s been put in charge of started as a spearhead project to discover the usability of Jira and over time the userbase grew until the whole company was using Jira for their day to day tasks and it’s now a core part of their BP.
As the first ever full-time Jira admin he has been tasked with resolving the issues users are having with slow loading times and bad responsiveness of the instance.
Using the Tables
Using the Task System
Reading the Audit log
Arima’s first step is to figure out what is causing the performance issues.
He looks at some of the performance metrics for the database and for the instance as a whole, everything points towards it being caused by the sheer amount of items on the instance.
To get help with cleaning up the abundance of items he installs Optimizer for Jira.
Arima starts out by going through the tables in Optimizer for Jira in the order described in a Jira cleanup article (this is going to link to the cleanup article).
Arima estimates the rough number of items that will be removed in each category and generates an excel spreadsheet of projects that are going to be deleted and messages the project leads for each to give them an opportunity to explain why the project shouldn’t be deleted.
After a two weeks the deadline for opting out of removal has passed, a database backup is scheduled and Arima adds all the projects to be deleted to the task queue in Optimizer in preparation. Once the backup completes Arima runs the deletion task and a notice is placed on the companies internal message board, informing users of the changes.
After a week without any complaints regarding the removal operations undertaken Arima proceeds to the next step of the cleanup procedure.
Arima goes through the tables and generates a detailed report of objects with no usages and their exclusive dependencies. Thanks to the removal of projects this list is quite a lot bigger compared to the original estimates.
A notice is placed on the companies internal message board with the list of objects that will be removed. All project leads are then instructed to go through the list and to inform Arima of any objects which should be excluded from the removal process along with a justification for the keeping of the objects.
After another two weeks the deadline passes and Arima finalizes his removal list.
A database backup and change freeze is scheduled for Friday evening.
Once Friday evening rolls around Arima goes through the tables and removes all the items in question in the order mentioned in the article, as this ensures dependencies are removed from the top down.
The cleanup operation is finished within the hour and a notice is once again placed on the internal message board.
Two weeks later Arima has finished configuring a RFC service desk where he has to approve the creation of any new configuration items or projects in Jira. A guide on how to request new items is placed on the companies Confluence site. Thanks to the new service desk the amount of new items added is reduced and with the added traceability the original cause of creation of any items can be easily assessed for future cleanup operations.