Why does flamegraph.pl filters the data? about flamegraph HOT 4 CLOSED

It doesn't filter. It will ignore invalid lines, and print that count as a warning. Sounds like your problem needs to be debugged!

from flamegraph.

Brendan, believe me, it does filter them. In order to reproduce here you are - two files:
https://drive.google.com/open?id=0B3DMXMfcPWF3T1ZSczdEM1U0N3c

the difference between sortedx.txt and sortedy.txt is in a dozen of lines: I filtered sortedx.txt and got sortedy.txt. Just run winmerge on these two files - and you will see that some lines are missing in sortedty.txt

So, if we run
./flamegraph.pl sortedy.txt > sortedy.svg
it must produce A LITTLE BIT LESS file than if we run
./flamegraph.pl sortedx.txt > sortedx.svg
right?

But in fact, I get sortedy.svg about 10MB in size, while sortedx.svg is 300 kb.
It should NOT be like this.

Note: I removed lines with these strings:

EPollArrayWrapper

org.apache.hadoop.net.unix.DomainSocketWatcher

java.lang.UNIXProcess.waitForProcessExit

org.apache.hadoop.util.Shell$1.run:509;java.io.BufferedReader.readLine:382;java.io.BufferedReader.readLine:317;java.io.BufferedReader.fill:154;java.io.InputStreamReader.read:184;sun.nio.cs.StreamDecoder.read:177;sun.nio.cs.StreamDecoder.implRead:325;sun.nio.cs.StreamDecoder.readBytes:283;java.io.BufferedInputStream.read:334;java.io.BufferedInputStream.read1:273;java.io.FileInputStream.read:272;java.io.FileInputStream.readBytes

org.apache.hadoop.util.Shell$1.run:509;java.io.BufferedReader.readLine:382;java.io.BufferedReader.readLine:317;java.io.BufferedReader.fill:154;java.io.InputStreamReader.read:184;sun.nio.cs.StreamDecoder.read:177;sun.nio.cs.StreamDecoder.implRead:325;sun.nio.cs.StreamDecoder.readBytes:283;java.io.BufferedInputStream.read:334;java.io.BufferedInputStream.read1:273;java.io.FileInputStream.read:272;java.io.FileInputStream.readBytes.:2

java.lang.Thread.run:745;java.util.concurrent.ThreadPoolExecutor$Worker.run:615;java.util.concurrent.ThreadPoolExecutor.runWorker:1145;java.util.concurrent.FutureTask.run:262;org.apache.hadoop.yarn.server.nodemanager.containermanager.launcher.ContainerLaunch.call:81;org.apache.hadoop.yarn.server.nodemanager.containermanager.launcher.ContainerLaunch.call:299;org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor.launchContainer:197;org.apache.hadoop.util.Shell$ShellCommandExecutor.execute:702;org.apache.hadoop.util.Shell.run:455;org.apache.hadoop.util.Shell.runCommand:524;org.apache.hadoop.util.Shell$ShellCommandExecutor.parseExecResult:715;java.io.BufferedReader.read:279;java.io.BufferedReader.read1:205;java.io.BufferedReader.fill:154;java.io.InputStreamReader.read:184;sun.nio.cs.StreamDecoder.read:177;sun.nio.cs.StreamDecoder.implRead:325;sun.nio.cs.StreamDecoder.readBytes:283;java.io.BufferedInputStream.read:334;java.io.BufferedInputStream.read1:273;java.io.FileInputStream.read:272;java.io.FileInputStream.readBytes.:2

These strings have too big quantity of calls (in comparison to the others). I think this is the reason of a bug.

from flamegraph.

It has minwidth for eliding data:

./flamegraph.pl --help
USAGE: ./flamegraph.pl [options] infile > outfile.svg

        --title       # change title text
        --width       # width of image (default 1200)
        --height      # height of each frame (default 16)
        --minwidth    # omit smaller functions (default 0.1 pixels)

from flamegraph.

so, if I want everything to be included, I need to specify --minwidth 0 and I will get everything.
THANK YOU VERY MUCH!

from flamegraph.