Using Dynatrace to replace Dareboost for frontend performance and quality audit?

[sean_kam_hung_c]

Hi all,

I would like to ask a question about using Dynatrace against Dareboost.

What we want is to see if we can fully leverage Dynatrace to replace Dareboost in terms of frontend performance monitoring and quality audit, but seems there are certain functionalities are currently missing with Dynatrace.

Therefore we would like to seek for advice if we miss out how Dynatrace can be fully utilised.

1. Page Score - from Dareboost there is an overall quality score provided to a site like https://www.dareboost.com/en/doc/discover-tool/quality-check. Is there any similar overall metric provided for quality audit?

2. Can Dynatrace perform quality audit on website frontend pages and suggest best practices to be followed on areas like browser rendering, quality SEO, security, accessibility?

3. Can Dynatrace setup automatic email alerts when issues are detected with the pages, e.g., from what mentioned on (2)?

4. In Dynatrace we can check frontend performance for a page but seems it is not instant? For example it would take at 15 minutes to get results from Dynatrace. Is it something we can still achieve to get the page performance in a more real-time manner?

Thanks.

Sean

[skrystosik · Accepted solution]

Hello,

1) each monitored application and user action in it has calculated APDEX index. It's done based on JS errors and performance.

2) It suggests findings, for example impact of 1st / 3rd party resources, CDN's, lacks in caching, resources sizes etc. But it will not tell you about SEO issues and security. At least not for now.

3) Dynatrace has AI engine called Davis. It is calculating all matrics and mades problem based on it. So if you have long time on some actions, failure rate, too high or too low traffic etc, Dynatrace will start problem. Based on configuration in DT problems can be sent to email, jira, service now, pagers duty etc + we have custom webhook integration. It allows you configure custom Webservice, for example SMS notifications.

4) It's not 15 mins, data from frontend are correlated with those that came from backend and this takes some time to calculate. In general it should be couple of minutes max (from my experience it's 5 min maximum, rather it's much quicker). I've never experienced times near to 15 minutes. The question is, how you are trying to check this time, because different dashboards have data in different time. For example user tag is detected in user session fast, but visit is tagged in user sessions after some period of time. But when you go inside session you will see this tag.

Sebastian

[suresh230591]

Hi,

In addition to what Sebastian has already mentioned for point 3 and 4

3. If you want to be alerted for JavaScript error and user action duration for certain critical actions, you can mark those actions as key user action and set individual thresholds.

4. I also never experienced such long delay in reporting the RUM data to DT GUI, It generally do not take more than 1-2 minute, worst case is 5 min.

[sean_kam_hung_c]

Hi Sebastian,

Thanks for your swift response. I would like to ask for further feedback based on follow-up questions as below:

• For question 3), can we understand if we can receive alert when we set a custom index, e.g., if page speed index is bigger than 5s? If yes is there any documentation available online to describe how to set this up?
• For question 4), actually 15mins means that we can only schedule the check / monitoring on periodic basis, e.g., 15 / 30 / 60 mins but we cannot conduct it instantly. Is my understanding correct?

Sean

[skrystosik · Accepted solution]

3) if you want alerting on performance of particular pages you have to mark them as key user actions. In such case you will have extra anomally detection settings for them and indeed you can make static thresholds for them.

4) I'm not sure what you are talking about. In general, problems will be raised instantly if condition will be met. Data in dynatrace as available as well almost instantly. What you can do is changing analysis time frame. Smallest period is 5 minutes, you can pick any timeframe you want. Data are not sent to dynatrace in chunks every X minutes, they are sent live.

Sebastian