Skip to main content
The info command provides a comprehensive, multi-section overview of a profile in a single command. It’s designed for quick triage and initial investigation.

​
Usage

ap-query info <file> [flags]

​
What It Shows

info combines multiple analysis views:
  1. Duration & samples - Recording span and sample count
  2. CPU vs WALL comparison - Thread group activity across event types (when both present)
  3. Thread distribution - Top threads by sample count
  4. Hot methods - Dual-ranked tables (self-time and total-time)
  5. Automatic drill-down - For top hot methods:
    • Call tree (callees)
    • Callers tree
    • Line number distribution
  6. Available events - Other event types in the file

​
Flags

​
--expand
int
default:"3"
Auto-expand top N hot methods with drill-down sections (tree, callers, lines). Set to 0 to disable drill-down.
​
--top-threads
int
default:"10"
Number of threads to show. Set to 0 to show all threads.
​
--top-methods
int
default:"20"
Number of hot methods to show in ranking tables. Set to 0 to show all methods.

​
Shared Flags

​
-e, --event
string
default:"cpu"
Event type: cpu, wall, alloc, lock, or hardware counter name.
​
-t, --thread
string
Filter to threads containing this substring.
​
--from
duration
Start of time window (JFR only). Examples: 5s, 1m30s, 2m.
​
--to
duration
End of time window (JFR only). Examples: 10s, 2m, 1m30s.
​
--no-idle
boolean
Remove idle leaf frames from analysis.

​
Examples

​
Basic Triage

ap-query info profile.jfr

Duration: 1m0.0s  Samples: 12043 (cpu)

=== CPU vs WALL ===
GROUP (threads)                   CPU%    WALL%
http-nio-exec (24)               78.3%   45.2%
Scheduled (4)                    12.1%   18.7%
GC Thread (8)                     8.2%    2.4%

=== THREADS (top 10) ===
http-nio-8080-exec-1             1824   15.1%
http-nio-8080-exec-2             1653   13.7%
ScheduledExecutor-1               892    7.4%
GC Thread#0                       543    4.5%
...

=== RANK BY SELF TIME (top 20) ===
METHOD                                              SELF%   TOTAL%   SAMPLES
HashMap.hash                                         8.3%   12.1%       998
ArrayList.grow                                       6.2%    6.2%       747
String.hashCode                                      5.1%    8.9%       614
...

=== RANK BY TOTAL TIME (top 20) ===
METHOD                                              SELF%   TOTAL%   SAMPLES
UserService.processRequest                           2.1%   45.3%      5456
HashMap.hash                                         8.3%   12.1%       998
...

Total samples: 12043

=== DRILL-DOWN: HashMap.hash (self=8.3%) ===
--- tree (callees) ---
HashMap.hash (total=12.1%, self=8.3%)
  HashMap.putVal (3.2%)
    HashMap.resize (2.1%)
  ...

--- callers ---
HashMap.hash (self=8.3%)
  HashMap.put (6.8%)
  HashMap.get (5.3%)
  ...

--- lines ---
HashMap.java:338         456    3.8%
HashMap.java:342         312    2.6%
...

Also available: wall, alloc

​
Focus on Specific Thread

ap-query info profile.jfr --thread worker

​
Minimal Drill-Down

ap-query info profile.jfr --expand 1 --top-methods 10
Shows only top 10 hot methods with drill-down for just the #1 hottest method.

​
Disable Drill-Down

ap-query info profile.jfr --expand 0
Skips all drill-down sections (tree, callers, lines).

​
Time Window Analysis

ap-query info profile.jfr --from 30s --to 1m
Analyze only samples between 30 seconds and 1 minute into the recording.

​
Output Sections

​
CPU vs WALL

Shown when both cpu and wall events are present. Groups threads by pattern and shows percentage in each event type.
High CPU% + High WALL% = CPU-bound work
Low CPU% + High WALL% = Waiting/blocking I/O
High CPU% + Low WALL% = Rare but possible with certain workloads

​
Threads

Lists individual threads by sample count. Thread names come from JFR metadata.

​
Hot Methods

Two tables ranked differently:
  • SELF TIME: Methods that appear at the top of stacks (doing actual work)
  • TOTAL TIME: Methods that appear anywhere in stacks (orchestrating work)

​
Drill-Down

Drill-down depth is fixed at 3 levels for tree and callers, 5 lines for line numbers. Use dedicated commands (tree, callers, lines) for deeper analysis.

​
When to Use

Initial Investigation

First look at a new profile to understand high-level patterns

Quick Triage

Rapidly assess performance characteristics without running multiple commands

Cross-Event Comparison

Compare CPU and wall-clock time to identify blocking vs compute

Thread Analysis

See which threads are active and what they’re doing

​
See Also

  • hot - Detailed hot method ranking with more options
  • threads - Thread analysis with grouping
  • tree - Deep call tree analysis
  • callers - Deep caller analysis

Build docs developers (and LLMs) love

Get started for freeTalk to us