07 Jun 2017 08:26 PM
Are there any tutorials for creating webchecks on sites that have dynamic content, or contstantly changing content? This would be especially useful for scripting against SaaS apps like Salesforce that have frequent changes made. I know it's possible to script against dynamic pages, and POST randomized content with the Advanced Synthetics script recorder, but I'm hoping to show this capability in Dynatrace Webchecks. Any suggestions on learning how to do this would be greatly appreciated.
Thank you,
Eric
Solved! Go to Solution.
08 Jun 2017 08:38 AM
Hi Eric,
is your question about breaking web check clickpaths caused by not working (outdated) locators? If yes, there is a setting to edit / fine-tune locators of a recorded clickpath.
Or is your question more about validating or providing dynamic content via web checks? It would be great if you can elaborate a little bit more on your use case or give a short example.
Kind regards, Philipp
08 Jun 2017 08:43 PM
Hey Philipp,
My question revolves around both scenarios:
Editing the locators of the clickpath seems to work great for constantly changing SaaS apps (like SalesForce) to script against, as long as you specifically identify the context of that particular CSS or DOM locator. I'm not very experienced in advanced scripting techniques, but I want to write one that can still click on a link or a button, regardless of the content of the link.
For example:
Is it possible to use the locators of a marketing banner that can be clicked, and when the banner changes, the script will continue to click on the new banner locator regardless of the actual banner content? I am hoping to incorporate change-resistant locators in my scripts.
The other question revolves around how to create a script that can accomplish both of following needs:
How do I accomplish this?
Regards,
Eric
09 Jun 2017 10:22 AM
Hi Eric,
thanks for the details. It should be possible to define locators that are flexible enough to cover your use case. However, it always depends on the individual page / application.
Random input - That's currently not possible with web checks but we have it on our roadmap.
Random click - Also not possible now. You could create multiple web checks to cover the most important scenarios but to be honest I don't think web checks are the right tool to test various combinations of a checkout process. That's maybe something that should be covered by functional UI tests beforehand.
On the other hand, additionally setting up real user monitoring for the checkout is crucial and would provide you with most important & meaningful data to find performance bottlenecks affecting your customers during the process.
Kind regards, Philipp
09 Jun 2017 01:21 PM
HI Philipp,
Thanks for the feedback. Yes, I agree thata RUM would be critical to understanding and resolving performance bottlenecks for some of these specific scenarios.
Thank You,
Eric