Initiating the Diagnosis of User-Experienced Client-Side Performance Issues
Hello community!
In this article (from my old blog), I want to divulge a simple technique I employ daily to tackle the frustrating sluggishness encountered by users when using Jira, Confluence, and the like.
I’ve also noticed this issue commonly crops up among end-users hesitant to transition to the Cloud, or those who face difficulties while accessing Atlassian products through VPN, and so on.
You might guess that I simply request users to submit a HAR file ( https://confluence.atlassian.com/kb/generating-har-files-and-analyzing-web-requests-720420612.html ).
Indeed, but asking users to produce a HAR file often leads to follow-up questions post-sharing a link or even a tutorial video. By then, the slowness might vanish, or issues like VPN disconnections or ISP troubles can arise. In such cases, I find offering a straightforward tool to pinpoint the issue can be more effective.
For several years now, I’ve been utilizing the “Page load time” extensions for both Google Chrome and Firefox.
Here’s the Chrome link: https://chrome.google.com/webstore/detail/page-load-time/fploionmjgeclbkemipmkogoaohcdbig
And the source code is available here: https://github.com/alex-vv/page-load-time
With this, I recommend users to compare the page loading times across various pages and sites before approaching me with issues. It’s always beneficial to have comprehensive and clearer logs when starting the diagnostic process.
Get Gonchik Tsymzhitov’s stories in your inbox
Join Medium for free to get updates from this writer.
For example, if it takes 4 seconds to load and render an Atlassian community page, I consider that pretty decent.
As for my cloud-hosted Atlassian Confluence, it loads significantly quicker, in just 1.32 seconds, which is impressive given my location in St. Petersburg, Russia.
Naturally, for internet-exposed installations, I would conduct an HTTP ping using different services like ping-host.
I am quite eager to hear about your methods too. How do you gather initial information for investigation?
While tools like New Relic provide insights through JavaScript monitoring, I aim to equip end-users with tools they can use themselves before coming back to me. Then, I usually request a HAR file and delve into the logs.
What’s your approach?
Best regards, Gonchik Tsymzhitov
