Microsoft MVP성태의 닷넷 이야기
.NET Framework: 2054. .NET Core/5+ SDK 설치 없이 dotnet-dump 사용하는 방법 [링크 복사], [링크+제목 복사],
조회: 15595
글쓴 사람
정성태 (techsharer at outlook.com)
홈페이지
첨부 파일
 
(연관된 글이 4개 있습니다.)
(시리즈 글이 6개 있습니다.)
Linux: 26. .NET Core 응용 프로그램을 위한 메모리 덤프 방법
; https://www.sysnet.pe.kr/2/0/12078

Linux: 27. linux - lldb를 이용한 .NET Core 응용 프로그램의 메모리 덤프 분석 방법
; https://www.sysnet.pe.kr/2/0/12083

.NET Framework: 2053. 리눅스 환경의 .NET Core 3/5+ 메모리 덤프를 분석하는 방법 - 두 번째 이야기
; https://www.sysnet.pe.kr/2/0/13135

.NET Framework: 2054. .NET Core/5+ SDK 설치 없이 dotnet-dump 사용하는 방법
; https://www.sysnet.pe.kr/2/0/13136

.NET Framework: 2055. 리눅스 환경의 .NET Core 3/5+ 메모리 덤프로부터 닷넷 모듈을 추출하는 방법
; https://www.sysnet.pe.kr/2/0/13137

.NET Framework: 2057. 리눅스 환경의 .NET Core 3/5+ 메모리 덤프로부터 모든 닷넷 모듈을 추출하는 방법
; https://www.sysnet.pe.kr/2/0/13139




.NET Core/5+ SDK 설치 없이 dotnet-dump 사용하는 방법

지난 글에 설명한 대로,

리눅스 환경의 .NET Core 3/5+ 메모리 덤프를 분석하는 방법 - 두 번째 이야기
; https://www.sysnet.pe.kr/2/0/13135

이제 덤프 파일 분석이 dotnet-dump의 개선으로 윈도우 환경에서만큼이나 쉬워졌습니다. 그런데, 여전히 문제가 하나 있습니다. 바로, dotnet-dump 자체가 (런타임이 아닌) SDK 설치를 요구하기 때문에,

# dotnet tool install -g dotnet-dump
It was not possible to find any installed .NET Core SDKs
Did you mean to run .NET Core SDK commands? Install a .NET Core SDK from:
    https://aka.ms/dotnet-download

가볍게 유지하려고 런타임만 설치된 docker 환경이나, 심지어 self-contained 유형으로 배포된 경우에는 설치 자체를 할 수 없다는 제약이 있습니다. 재미있는 건, 사실 dotnet-dump가 SDK 설치 없이도 잘 동작한다는 점입니다. 그래서 그냥 dotnet-dump가 설치된 리눅스 환경에서 복사를 해올 수 있습니다. 방법도 간단합니다. ^^

일반적으로 dotnet-dump는 현재 사용자 계정의 "./.dotnet/tools" 밑에 생성이 되고,

$ which dotnet-dump
/home/testusr/.dotnet/tools/dotnet-dump

그리고 dotnet-dump 실행 파일 기준으로 ".store" 디렉터리 하위에 dotnet-dump의 설치 버전이 요구하는 부가적인 실행 모듈들이 위치하게 됩니다. 따라서 이 구조만 지켜서 다음과 같이 (약 8MB 정도의) 압축 파일로 만들어 주면,

dotnet-dump-self_1.png

이것을 .NET Core/5+ SDK가 설치되지 않은 환경, 예를 들어 아래는 "mcr.microsoft.com/dotnet/core/aspnet:3.1-buster-slim" 이미지를 기반으로 한 docker 컨테이너에 복사해 실행한 예를 보여줍니다.

# apt install unzip

# unzip dotnet-dump-exec.zip

# chmod 744 dotnet-dump

# ./dotnet-dump ps
 1  dotnet  /usr/share/dotnet/dotnet  dotnet razor31_sample.dll  

# ./dotnet-dump collect -p 1

Writing full to /app/temp/core_20221007_093929
Complete

보다시피 정상적으로 덤프를 완료했습니다. 이제 저 파일을 다른 PC에 복사해 "리눅스 환경의 .NET Core 3/5+ 메모리 덤프를 분석하는 방법 - 두 번째 이야기" 글에서 설명한 것처럼 편하게 분석하시면 됩니다. ^^

대신, 분석(analyze) 기능은 .NET 런타임 또는 SDK가 설치돼 있어야 사용할 수 있습니다. 가령, 닷넷 런타임이 없는 환경에서 analyze 옵션으로 실행하면 이런 오류가 발생합니다.

$ dotnet-dump analyze core_20221007_093929 
A fatal error occurred. The required library libhostfxr.so could not be found.
If this is a self-contained application, that library should exist in [/home/testusr/temp/.store/dotnet-dump/6.0.328102/dotnet-dump/6.0.328102/tools/netcoreapp3.1/any/].
If this is a framework-dependent application, install the runtime in the global location [/usr/share/dotnet] or use the DOTNET_ROOT environment variable to specify the runtime location or register the runtime location in [/etc/dotnet/install_location].

The .NET runtime can be found at:
  - https://aka.ms/dotnet-core-applaunch?missing_runtime=true&arch=x64&rid=ubuntu.18.04-x64&apphost_version=5.0.17

참고로, 리눅스 환경에서 뜬 덤프일지라도 윈도우 환경의 dotnet-dump로 분석할 수 있습니다.




물론, gcore/gdb/procdump 등을 이용해 덤프를 뜨는 방법도 있습니다. 하지만, 이것을 이용해 덤프를 뜨고 싶어도 대개의 일반적인 운영 환경에서는 해당 모듈을 설치해야 합니다. 차라리 8MB 크기의 dotnet-dump zip 파일을 복사하는 것이 낫습니다.

게다가 아래의 문서를 보면,

Dump collection and analysis utility (dotnet-dump)
; https://github.com/dotnet/diagnostics/blob/main/documentation/dotnet-dump-instructions.md

"dotnet-dump analyze" 명령어는 오직 "dotnet-dump collect"로 수집한 덤프에 대해서만 분석을 지원하고 있는 듯합니다. 따라서, gcore 기반의 덤프를 분석하려면 lldb를 이용해야 하는데요, 이를 위해서는 아래의 문서에서 설명한 준비 작업이 필요합니다.

Lab 1.2 Troubleshooting crashes - analyze system-generated core dump files in lldb debugger
; https://learn.microsoft.com/en-us/troubleshoot/developer/webapps/aspnetcore/practice-troubleshoot-linux/lab-1-2-analyze-core-dumps-lldb-debugger

// 2022-10-08 업데이트: 아래의 문서를 보면 gcore로 뜬 덤프도 dotnet-dump에서 분석할 수 있다고 합니다.
Lab 4.2 Analyze core dump files on another machine - Using WSL to open core dump files
; https://learn.microsoft.com/en-us/troubleshoot/developer/webapps/aspnetcore/practice-troubleshoot-linux/lab-4-2-analyze-core-dumps-another-machine-wsl

역시나 차라리 저런 준비를 하느니 dotnet-dump zip 파일을 복사해 덤프를 뜨는 것이 훨씬 간편합니다.

그리고 결정적으로, gcore를 기반으로 덤프를 뜨면 (비주얼 스튜디오 예제 웹 프로젝트라도) 수십 GB의 덤프 파일이 생성될 수 있습니다.

# pmap 1 | grep total
 total         63416700K // pmap을 이용해 대상 프로세스의 덤프 파일 크기를 미리 예측

# apt-get install gdb zip
# gdb

(gdb) attach 1

(gdb) generate-core-file /app/core_31.dmp
warning: target file /proc/1/cmdline contained unexpected null characters
Saved corefile /app/core_31.dmp
(gdb) quit

# zip core_31.zip core_31.dmp
  adding: core_31.dmp (deflated 100%)

# ls -l core_31.zip
-rw-r--r-- 1 root root 109822428 Oct  6 21:53 core_31.zip

제가 테스트한 "mcr.microsoft.com/dotnet/core/aspnet:3.1-buster-slim" 컨테이너의 경우 59GB라는 어마어마한 덤프 파일을 생성했는데요, 운영 서버라면 그 정도 크기의 덤프 파일을 생성할 때까지 서비스가 중지된다는 것을 의미합니다. (같은 상황에서 dotnet-dump는 2GB 파일의 덤프를 생성합니다.)

그러니까, 다른 거 생각하지 마시고... 그냥 무조건 닷넷 덤프는 "dotnet-dump"로 처리하는 걸로 결정하시면 되겠습니다. ^^




참고로, 그냥 귀찮다면 ^^ dockerfile에 sdk 설치하는 방식으로 구성할 수 있습니다.

# SDK 버전 3.1을 가정

RUN apt-get install wget -y
RUN wget https://dot.net/v1/dotnet-install.sh -O dotnet-install.sh
RUN chmod +x dotnet-install.sh
RUN ./dotnet-install.sh --channel 3.1

RUN /root/.dotnet/dotnet tool install -g dotnet-dump --version 3.1
RUN ln -s /root/.dotnet/tools/dotnet-dump /usr/bin/dotnet-dump




[이 글에 대해서 여러분들과 의견을 공유하고 싶습니다. 틀리거나 미흡한 부분 또는 의문 사항이 있으시면 언제든 댓글 남겨주십시오.]

[연관 글]






[최초 등록일: ]
[최종 수정일: 3/31/2025]

Creative Commons License
이 저작물은 크리에이티브 커먼즈 코리아 저작자표시-비영리-변경금지 2.0 대한민국 라이센스에 따라 이용하실 수 있습니다.
by SeongTae Jeong, mailto:techsharer at outlook.com

비밀번호

댓글 작성자
 




[1]  2  3  4  5  6  7  8  9  10  11  12  13  14  15  ...
NoWriterDateCnt.TitleFile(s)
13945정성태6/7/2025427오류 유형: 960. 파이썬 + conda - mysqlclient 사용 시 "NameError: name '_mysql' is not defined" 에러
13944정성태6/7/2025462오류 유형: 959. The trust relationship between this workstation and the primary domain failed. - 네 번째 이야기
13943정성태6/6/2025653개발 환경 구성: 748. Windows + Foundry Local - 로컬에서 AI 모델 활용
13942정성태6/5/2025862오류 유형: 958. winget 설치 시 "0x80d02002 : unknown error"
13941정성태6/2/20251024닷넷: 2334. C# - cpuid 명령어를 이용한 CPU 제조사 문자열 가져오기파일 다운로드1
13940정성태6/1/20251406C/C++: 188. C++의 32비트 + Release 어셈블리 코드를 .NET으로 포팅할 때 주의할 점파일 다운로드1
13939정성태5/29/20251696오류 유형: 957. NVIDIA Triton Inference Server - version `GLIBCXX_3.4.32' not found (required by /opt/tritonserver/backends/python/triton_python_backend_stub)
13938정성태5/29/20251435개발 환경 구성: 747. 파이썬 - WSL/docker에 구성한 Triton 예제 개발 환경
13937정성태5/24/20251356개발 환경 구성: 746. Windows + WSL2 환경에서 (tensorflow 등의) NVIDIA GPU 인식
13936정성태5/23/20251188개발 환경 구성: 745. Linux / WSL 환경에 Miniconda 설치하기
13935정성태5/20/20251231파이썬 - pip 사용 시 "ImportError: cannot import name 'html5lib' from 'pip._vendor'" 오류
13934정성태5/20/20251709스크립트: 77. 파이썬 - 'urllib.request' 모듈의 명시적/암시적 로딩 차이
13933정성태5/19/20251288오류 유형: 956. Visual Studio 2022가 17.12 버전부터 업데이트 되지 않는다면?
13932정성태5/18/20251499스크립트: 76. 파이썬 - Version 문자열 다루기(semver 패키지)
13931정성태5/17/20251791스크립트: 75. 파이썬 - Cython 기본 예제 및 컴파일
13930정성태5/17/20251489개발 환경 구성: 744. 파이썬 - Windows embeddable package 환경에서 외부 패키지 사용하는 방법(ex: UFO² 환경 구성)
13929정성태5/16/20251509오류 유형: 955. 파이썬 - "Windows embeddable package" REPL 환경에서 "NameError: name 'exit' is not defined"
13928정성태5/15/20251553오류 유형: 954. UFO² - "'Invalid URL (POST /v1/chat/completions/chat/completions)'"
13927정성태5/15/20251545오류 유형: 953. OpenAI - The API request of HOST_AGENT failed: OpenAI API request exceeded rate limit: Error code: 429
13926정성태5/14/20251907개발 환경 구성: 743. LLM과 윈도우의 만남 - Desktop AgentOS UFO² 기본 환경 구성
13925정성태5/12/20252006닷넷: 2333. C# - (Console 유형의 프로젝트에서) Clipboard 연동파일 다운로드1
13924정성태5/8/20251757닷넷: 2332. C# - (JetBrains Omea Reader 대상으로) 런타임 시에 메서드 가로채기 [2]파일 다운로드1
13923정성태5/5/20251503스크립트: 74. 파이썬 - C# - Python.NET의 RunSimpleScript, Exec, Eval 차이점파일 다운로드1
13922정성태5/3/20251751스크립트: 73. 파이썬 - Windows embeddable package 버전에서 tkinter 환경 구성
13921정성태5/3/20252265오류 유형: 952. 듀얼 채널 메모리 정렬을 지키지 않은 컴퓨터의 Windows 비정상 종료 현상(Blue Screen) [2]
[1]  2  3  4  5  6  7  8  9  10  11  12  13  14  15  ...