You support a high-traffic web application that runs on Google Cloud Platform (GCP). You need to measure application reliability from a user perspective without making any engineering changes to it. What should you do? (Choose two.)
A.
Review current application metrics and add new ones as needed.
B.
Modify the code to capture additional information for user interaction.
C.
Analyze the web proxy logs only and capture response time of each request.
D.
Create new synthetic clients to simulate a user journey using the application.
E.
Use current and historic Request Logs to trace customer interaction with the application.
A&D. Why E isnt correct - Synthetic transaction already provides the capability mentioned in E - "Use current and historic Request Logs to trace customer interaction with the application". Instead, option A - to review and add additional metrics makes sense !!
It says "measure application reliability from a user perspective" - you need synthetic measurements from outside the infrastructure (D). To simulate the synthetic journey mentioned, you first need to understand the current user behaviours using the 'request' logs (E).
Web proxy is not a reverse proxy, it is a forward proxy - a type of server that runs at the client side. If you have clients from all over the world, how will you collect their proxy logs?
So it cannot be C - D&E should be the answer which don't require any engineering changes "in the" application.
A - Adding metrics doesn't necessarily reflect reliability from a user perspective'
B - Modifying the code is against the requirement
C - Analyzing proxy logs doesn't connect the findings to a user perspective
From the request logs, user journeys can be recreated (E) and the replays can be fed to synthetic clients to simulate/replay (D).
1.Based on " without making any engineering changes to it" , exclude A,B at first.
2. Based on following described,
https://cloud.google.com/architecture/adopting-slos#choosing_a_measurement_method
D & E should be better.
Answer: CE
A. Review current application metrics and add new ones as needed.
==> (x) "add new ones" needs engineering changes
B. Modify the code to capture additional information for user interaction.
==> (x) "Modify the code" needs engineering changes
C. Analyze the web proxy logs only and capture response time of each request.
==> (O) no engineering changes
D. Create new synthetic clients to simulate a user journey using the application.
=> (x) "Create new synthetic client" needs engineering changes.
E. Use current and historic Request Logs to trace customer interaction with the application. ==> (O) no engineering changes
This two option doesn't require engineering changes into the application. Web Proxy logs is a forward proxy thing so it present in client side. others need changes
C& E.
C-> a web proxy relays URL requests from clients to a server. Analyzing web proxy logs can give unobtrusive insights into the browsing behavior of users
A&D should be the correct answer:
D - Easy/default choice since this doesnt need any changes on the app
A - Since the ask is to "measure reliability" - review and add the appropriate SLIs. Option E -tracing is needed only for collecting/troubleshooting latency issues - hence not the correct choice
To me , it looks C & E
B & D, need engineering changes and investment, hence I ruled these two out
This leaves us with A, C and E - The question is asking for "Need to measure Application Reliability from User perspective without Engineering Changes" - This rules out A, as talks about adding new metric but not stating which one
Considering high-traffic web application as "Request Driven Services" - Two of the suggested SLI's are Availability and Latency - C & E to me covers Latency
https://cloud.google.com/architecture/adopting-slos?hl=en
umm but
> C. Analyze the web proxy logs 'only' and
the phrase 'only' bothers me...
we can know the reliability by getting the status of web applications by web proxy logs(I mean HTTP status is good/bad), but we don't need to view these logs 'only'...
A voting comment increases the vote count for the chosen answer by one.
Upvoting a comment with a selected answer will also increase the vote count towards that answer by one.
So if you see a comment that you already agree with, you can upvote it instead of posting a new comment.
ralf_cc
Highly Voted 2 years, 10 months agoBiden
2 years, 4 months agoRupo7
1 year, 9 months agoCharun
Highly Voted 2 years, 10 months agocetanx
2 years, 7 months agojomonkp
Most Recent 4 months, 4 weeks agoenter_co
1 year, 5 months ago[Removed]
1 year, 11 months ago[Removed]
1 year, 10 months agopaul223344
2 years agoPhilipKoku
2 years, 2 months agocloudbee
2 years, 3 months agoguid1984
2 years, 4 months agorspstudent
2 years, 4 months agoFreemiums
2 years, 4 months agonot_thanos
2 years, 4 months agoBiden
2 years, 5 months agoraf2121
2 years, 8 months agogcpengineer
2 years, 8 months agokubosuke
2 years, 9 months ago[Removed]
2 years, 7 months agokubosuke
2 years, 9 months agogcpengineer
2 years, 8 months agoakg001
2 years, 10 months agoholahola
2 years, 10 months agorinkeshgala1
2 years, 10 months ago