We have identified Jira Dashboards can be a potential bottleneck while helping our customer to tune the performance of their Jira. This article explains why Jira takes longer time and give some tips on how to speed things up.
Why my Jira Dashboard takes a long time to load?
These are some factors that contributes to the slowness of Jira:
- Huge number of gadgets within a single dashboard
- Gadgets with complex reporting
- Filters with a huge number of issues
Huge number of gadgets
Whenever a Jira Dashboard page is loaded, the browser will send a number of requests to the server for all the CSS and Javascript required. (For more details, check out JRASERVER-62126). When there are more gadgets, it will fire more requests.
It will be faster if you focused doing 1 task at a time versus doing 100 tasks concurrently. By the same principle, your dashboard will load faster if Jira has less requests to work on at the same time.
Jira helps with a default maximum limit of 20 gadgets within a dashboard. However, it is possible to modify the limits on the number of gadgets on a dashboard.
When there are way too many requests
We have a couple of support tickets which the end users added a lot of Gauge Gadgets in their dashboard. As a result, some gadgets on the dashboards cannot load.
This is because the browser will silently throw the error message Failed to load resource: net::ERR_INSUFFICIENT_RESOURCES if it detects that are way too many requests within a short interval.
To support the users, we introduced Multiple Filters Counter Gadget that can display multiple counter within a single gadget.
Complex Reporting Gadgets
Not all Jira gadgets are the same. There are some gadgets which involves complex processing. For example, our Tissue for Jira app performs the handy task of traversing all the linked issues and extracting the various field values to present a tabular overview.
If you are using such complex reporting gadget, it will be advisable to have lesser gadgets within that dashboard.
Filters with many matching issues
It is likely that some dashboards load very fast when they were created initially. However, as the number of issues in the project increases over time, the performance of the dashboard become slower without the original author noticing.
This is because there are more issues to be processed. A useful tip will be to time-box your Jira filters within a time period using JQL date functions like startOfYear(), startOfMonth()
Alternatively, you can use our Rolling Window Monthly/Weekly Gadgets which only retrieve the matching issues in the last X weeks/months specified.
Other ways to load your Jira Dashboards faster
Add more computing power
The simplest way is to pump more computing resources like more CPU and memory. Jira Data Center also scales the performance by distributing the workload across more nodes.
Split into multiple dashboards
As mentioned previously in Best Practices in Jira Dashboard Reporting, it is recommended to keep a dashboard to its objectives to allow people to identify the action required.
But it is troublesome to have many dashboards
Beside adding the links to various dashboards as project shortcuts in your Jira project, you can also add links to related dashboards using our free Link Menu Gadget to facilitate navigation. You can also add links to the your Confluence spaces and other related project resources too.
If that is still not enough and you want to access your dashboards easily from everywhere in Jira. You can organise your Dashboards in cascading Dashboard Folders which can be accessible in the Jira top menu.
What is the performance of your Jira Dashboard?
You may want to do the 23 seconds test on your frequently used dashboards. If it is taking longer than that, you might want to tidy up your dashboards.
