In this article, I want to share my experience to view Presto trace using watsonx.data 2.1.3 with Instana. After you configure OpenTelemetry in watsonx.data using Instana as the backend, next step is to view telemetry data using Instana User Interface.

This article has next topics.

• Keywords to filter Presto trace
  The following 2 sub-sections are for Instana beginner.
  • 1st Step to filter Presto trace
  • Set filter using "Add filter"
• Details of Presto trace
• Details of Presto trace when an error occurs

If you don't configure it yet, my blog watsonx.data OpenTelemetry : Overview and Configuration will help you.

watsonx.data OpenTelemetry Overview

In watsonx.data 2.1.3 and 2.2.0, OpenTelemetry is supported for Presto (Java) engine and Milvus service.
Trace in OpenTelemetry represents the lifecycle of a single operation or request as it propagates through a system, capturing spans to detail its execution across services.

Keywords to filter Presto trace

• Service Name is "Presto".
• When a SELECT query is invoked, a "POST /v1/statement" trace is generated. Call Name is "POST /v1/statement".

The following sub-sections are for Instana beginners (like me).

1st Step to filter Presto trace

1. In the Instana web user interface, select Analytics.
   
2. You get panel "Analytics > Application/Calls" by the default. Select "Presto" at the left side, only Presto traces are listed.
   
   You can set filter using Latency the unit is millisecond. You can also set a time range.
   

Set filter using "Add filter"

Here, I show you the way to filter using "Add filter".

1. Add filter "Service > Name = Presto" : Click Add filter and scroll down and select "Name" under "Service". Select "=" and input "Presto"
   

2. When a SELECT query is issued, a "POST /v1/statement" trace is generated.　To add, click Add filter and add "Call > Call Name = POST /v1/statement".
   

3. In case you have multiple OCP clusters with watsonx.data, you can filter using "ClusterName". In my case, it is "api.itz-xczjzg.infra0xxxxxxxx" which is hostname part in OCP cluster API URL. Add filter "OpenTelemetry Tags > ClusterName equals api.itz-xczjzg.infra0xxxxxxxx".
   (1) Click "Add filter", select "OpenTelemetry Tags" under "OpenTelemetry", and input "ClusterName" after it.
   (2) Select "equals" and input "ClusterName" like "api.itz-xczjzg.infra0xxxxxxxx"
   

4. You can use Latency in Call or Trace. Set the value in milliseconds.
   
   

Details of Presto trace

After you narrow down the Presto traces, you will find a trace to drill down details.

Here, I want focus on "POST /v1/statement" which takes 27,973ms. Next is the Trace details.
Scroll through the view as seen on the left to get to the view as seen on the right.

The left and right sides represent the same calls. The left side is a visual representation of the time required, and the right side is a Tree display that makes it easy to see the progress of the process.

• Note: "Duration" is shown 9.29 min in this screen shot, but don't worry. The last DELETE call comes out after 9 minutes. The actual time required for Query is expressed in "query" call and is 27,973ms.

Next, I want to focus on a "query" call. When you click on each call, the details of that call will appear on the right side in Instana UI.
Click on "query" call, you will get the view as seen in screen shot below.

Because almost all items are displayed with truncated, I added whole text at the right.

• The "query" section has Started(timestamp), Latency, etc.
• The "Details" section has Service "Presto".
• The "Tags" section has items directly related to watsonx.data or Presto engine.
  • ClusterName : hostname part in OCP cluster API URL
  • Query : full text of query
  • Instanceid : watsonx.data instance ID

Details of Presto trace when an error occurs

Next screenshot is an example of "POST /v1/statement" which reports an error.

By clicking error call, you can get error details, like ERROR_NAME, ERROR_CODE, Query text.
With "download" icon, more detailed error information is downloaded.

Conclusion

This article shows filtering Presto trace in Instana User Interface and examples of Presto trace.

Environment

The example in this topic introduced is run in the following environment.

• OCP Version : 4.16
• CP4D 5.1.3 , watsonx.data 2.1.3
  • Presto engines
    • Presto (Java) v0.286 / Size : Starter
    • Presto (C++) v0.286 / Size : Starter
  • Milvus service : Version v2.5.0 / Size : Small
• Instana (June 2025)