Competitive Analysis: Logging
What’s this issue all about?
The purpose of this issue is to conduct the competitive analysis in the Logging space to better understand the desired user experience for Logging, based on what competitor solutions are offering.
Who is the target user of the feature?
What questions are you trying to answer?
The competitive analysis will help us answer the following questions:
- What is the sorting order for logs offered by our competitors? Is the most recent log entry at the top or at the bottom of the UI?
- How are the logs displayed? (table view, etc.) What are pros and cons of different ways to display the logs?
- What's the information that our competitors show within the logs?
- What are the UX patterns used to navigate to a specific place in the logs? Are our competitors using date ranges on single timestamp navigation?
- What are the UX patterns used to navigate through pages of logs (for ex. pagination or infinite scroll)? What are the pros and cons of each approach?
- What actions can be performed with the information in the logs?
- Where can the terminated pod logs be accessed/ filtered by, and how are they displayed?
- What are the ways to filter the pods list (for ex. all pods vs. active pods vs. terminated pods)?
Competitors we'll be testing
How are our main competitors solving the problems we're aiming to solve in the upcoming iterations? (see the questions above)
- Elastic
- Honeycomb
- Splunk
- Datadog
- Sumologic
Process we'll follow
- Conduct the testing of competitor products focusing on the specific questions outlined above.
- Record the testing sessions on video and link the videos in this issue description to make the insights about the product interactions accessible to the rest of the team.
- Collect the insights in the Mural as post-it notes, accompanied by screenshots to illustrate the insights.
- Conduct analysis of the research results, looking for patterns and core insights. For example, a core insight may be "5 out of 6 competitor products allow the user to filter the logs by pod status".
- Add the core insights to the UXR Insights repository.
- Share the learning from this effort and the process we used with the rest of the team.
Steps for performing a testing session
-
Start your screen recorder -
Look at the first question to be compared -
View competitor A and look for the feature referenced in the question -
Take a screen cap and past it into Mural under the appropriate company and question section -
Test out the feature in question, narrating your experience, taking notes in Mural, and taking screen caps where necessary along the way -
Repeat the steps for all questions/ features -
Once you're done, finish the recording, upload to GDrive, and paste the link into Mural
What hypotheses and/or assumptions do you have?
What decisions will you make based on the research findings?
Inform the next iterations of our Logging solution.
What's the latest milestone that the research will still be useful to you?
13.0
Additional links
Finding
- We should consider changing our display and show the latest logs at the top, (some vendors provide the ability to sort the time )column
- All vendors provide log histogram
- All vendors offer query bar with a supported query language for log exploration - we should leverage our filtered search and offer similar filtering capabilities (e.g, and/or/not operations for logs fields that Elastic provides)
- Logs are displayed in a table view the values in the table are clickable and populate a query bar with the right query (e.g. clicking a host name will filter all logs with host name =value)
- Logs are displayed in the following order: Time and text, where possible additional metadata is added displayed in the left-hand side of the table (e.g. such as log name, source/collector, pod name/id, docker name/id)
- Navigation to and from the log explorer is possible from different charts (e.g. metrics to logs)
- When selecting a time picker it is normally a range and not a single time stamp all vendors offer both quick timerange (last X min) or customized date and time range
- The information is clickable and populate the query bar with the query (e.g. clicking on a value in the table added this value to the search criteria)
- Pod logs are available in the same fashion as other logs, I wasn't able to find a vendor that specifically provides a method to query for terminated pod logs, however, this is doable using the right query, the need for looking at terminated pod logs should be fleshed out using customer interviews for further validation however I believe that this might be a potential differentiator in the log explorer
Edited by Nadia Sotnikova