Stack Overflow
Stack Overflow

Questions tagged [flamegraph]

Flame graphs are a visualization of profiled software, allowing the most frequent code-paths to be identified quickly and accurately — Brendan Gregg

Flame graphs are a visualization of profiled software, allowing the most frequent code-paths to be identified quickly and accurately.

enter image description here

The x-axis shows the stack profile population, sorted alphabetically (it is not the passage of time), and the y-axis shows stack depth. Each rectangle represents a stack frame. The wider a frame is, the more often it was present in the stacks. The top edge shows what is on-CPU, and beneath it is its ancestry. The colors are usually not significant, picked randomly to differentiate frames.

This visualization is fully introduced and explained in the CPU Flame Graphs page, and in Brendan Gregg presentation.

63 questions
1
vote
1 answer

Construct flamegraph with start and end timestamps

i am collecting profile with custom instrumentation of each function and the format is simple. Each line in the profile dump has the following data: symbol address, start time, end time Is it possible to convert this to a format consumable by…
A. K.
  • 34,395
  • 15
  • 52
  • 89
1
vote
0 answers

Creating flamegraphs over specific binary executable

I want to create a flamegraph of my Rust executable (which is in target/release/name-of-binary). I found the perf command to here However, since I'm suing mac os, I don't have perf and have to use dtrace. Although I figured out how to use dtrace for…
Ziva
  • 3,181
  • 15
  • 48
  • 80
1
vote
0 answers

Flamegraph shows a lot of overhead from perf_event

I am new to BPF and perf, I am using perf record -F 150 -g -p $(pidof app_name) -- sleep 60 to profiling a pub sub based app, the publisher side is around 100HZ. and flamegraph is dominated by perf_event, which I think it is related to the…
baltam
  • 53
  • 5
1
vote
0 answers

Running cargo flamegraph from bash works, fails using process::Command from Rust

I have been using the flamegraph profiler, rs-flamegraph, quite successfully over the past few days. When I run it from bash it works fine. No sudo or anything. ~/.cargo/bin/flamegraph -o "$f flamegraph.svg" ./binary When I invoke the binary from…
ArekBulski
  • 4,520
  • 4
  • 39
  • 61
1
vote
0 answers

Clinic flamegraph: trying to read graph to determine fixes

I used Clinic to create a flamegraph and I am seeing a lot of activity coming from Mongoose. A common theme seems to indicate that a lot of these calls seems to be "line: 1, column:1". On the surface, It seems like there are ways to optimize…
user1790300
  • 2,143
  • 10
  • 54
  • 123
1
vote
0 answers

How to include specified function in the flame graph?

There is no function name that I want to see in the flame graph, but I'm sure this function works. How can I do to show the function name what I want to see displayed in the flame graph?
rocco
  • 21
  • 4
1
vote
1 answer

Comiling linux perf tools and its dependent kernel modules from source/scratch

I am using linux perf tool for profiling shared library. Though it worked well on Ubuntu but now I want to run it on embedded linux and I cannot use apt-get to install linux perf tools on embedded linux. That's why I should have to compile…
Arslan Ali
  • 63
  • 1
  • 12
1
vote
0 answers

How to analyze flame graphs for web application

I have brief understanding on sampling concept and watched some flame graphs videos. Started jump in to do performation analsis for an e-commerces web application(which is powered by Odoo Monolic, Python with Werkzeug WSGI server), I have some…
Veerendra K
  • 2,145
  • 7
  • 32
  • 61
1
vote
1 answer

Detecting the cause of 15% cpu usage in java web application

Recently we upgraded hibernate jars in our java web application. After this upgrade we found 15-20% increase in the cpu usage. The only difference before and after is the hibernate jar version. I need to determine the root cause of this increase in…
Andy Dufresne
  • 6,022
  • 7
  • 63
  • 113
1
vote
1 answer

profiling java server and interpreting flame graph

I am profiling java server . I am looking at the flame graph : http://www.brendangregg.com/flamegraphs.html . I have couple of questions on how to interpret flame profile . I am new to java, so they may sound very basic. 1) in the profile , the top…
user179156
  • 841
  • 9
  • 31
1
vote
1 answer

Dask dashboard profile tab usage (a.k.a. Flame graph)

How do you interpret this piece of tetris to study the usage of your application?
matanster
  • 15,072
  • 19
  • 88
  • 167
1
vote
1 answer

Reading the output of profvis in R

I am trying to learn to use profvis. Here is a link with a reproducible example: profvis example I am not able to see why there are TWO numbers for Memory ie. -3.5 and 9.5 (where Time = 1290) in the first example in this page. Can someone explain…
user2338823
  • 501
  • 1
  • 3
  • 16
1
vote
0 answers

How do I get detailed flamegraphs with the flame crate for code written using Rayon?

I'm trying to get some performance metrics using the flame crate with code I've written using Rayon: extern crate flame; flame::start("TAG-A"); //Assume vec is a Vec vec.par_iter_mut().filter(|a| a == 1).for_each(|b| func(b)); //func(b)…
Darakian
  • 619
  • 2
  • 9
  • 22
1
vote
3 answers

Perf Tool Linux - No Call-Graph

I am tring to use perf version 3.0.8 to profile my embedded system Linux kernel version 3.0.8. I cross-compiled tools/perf using zlib and elfutils. In order to familiarize with this tool I am using a simple quicksort example compiled using -g…
Mauro P.
  • 21
  • 5
1
vote
1 answer

Dealing with information overload for rack mini profiler flamegraph

The flamegraph created by rack mini profiler has a lot of information, but I'm having difficulty dealing with it all, because of the large number of layers involved: Is it possible to hide the layers involving third-party gems, or display the…
Andrew Grimm
  • 78,473
  • 57
  • 200
  • 338
1 2
3
4 5