[B! perf] yassã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

yass id:yass

perfã«é–¢ã™ã‚‹yassã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (19)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

perf + Flame Graphs ã§ Linux ã‚«ãƒ¼ãƒãƒ«å†…ã®ãƒœãƒˆãƒ«ãƒãƒƒã‚¯ã‚’ç‰¹å®šã™ã‚‹ - ablog
Linuxã§ddã§1GBã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚’ä½œæˆã— perf ã§ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚°ã—ã€Flame Graph (ç‚Žã®ã‚°ãƒ©ãƒ•ï¼Ÿ)ã«ã—ã¦å¯è¦–åŒ–ã—ãŸã‚‚ã®ã§ã™ã€‚ Flame Graphs ã¯ perf(Linux)ã€SystemTap(Linux)ã€DTrace(Solarisã€Oracle Linux(UEK)ã€Mac OS Xã€FreeBSD)ã€XPerf.exe(Windows) ãªã©ã§ã®ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚°çµæžœã‚’å¯è¦–åŒ–ã—ã¦æœ€ã‚‚ä½¿ã‚ã‚Œã¦ã„ã‚‹ã‚³ãƒ¼ãƒ‰ãƒ‘ã‚¹ã‚’æ—©ãæ£ç¢ºã«ç‰¹å®šã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚å®Ÿä½“ã¯ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚°çµæžœã‚’ã‚°ãƒ©ãƒ•(SVG)ã«å¤‰æ›ã™ã‚‹ Perl ã‚¹ã‚¯ãƒªãƒ—ãƒˆã§ã™ã€‚ ä¸‹ã‹ã‚‰ä¸Šã«è¡Œãã»ã©ã‚³ãƒ¼ãƒ«ã‚¹ã‚¿ãƒƒã‚¯ãŒæ·±ãã€å·¦ã‹ã‚‰é–¢æ•°åã®ã‚¢ãƒ«ãƒ•ã‚¡ãƒ™ãƒƒãƒˆé †ã§ã‚½ãƒ¼ãƒˆã•ã‚Œã¦ã„ã¾ã™ã€‚ä¸€ç•ªä¸Šã§æ¨ªå¹…ãŒåºƒã„é–¢æ•°ãŒCPUã‚’é•·ãä½¿ã£ã¦ã„ã¾ã™ã€‚ä»Šå›žã¯ "_aesni_enc1" ã¤ã¾ã‚Šæš—å·åŒ–ãŒãƒœãƒˆãƒ«ãƒãƒƒã‚¯ã«ãªã£ã¦ã„ã‚‹ã“ã¨ãŒã‚ã‹ã‚Šã¾ã™ã€‚ ã‚·ã‚¹ãƒ†
yass 2015/07/07
perf

flame graphs
ãƒªãƒ³ã‚¯
brendangregg/perf-map-agent Â· GitHub - A java agent to generate method mappings to use with the linux `perf` tool
yass 2015/02/08
java

perf
ãƒªãƒ³ã‚¯
brendangregg/perf-tools Â· GitHub - Performance analysis tools based on Linux perf_events (aka perf) and ftrace
A miscellaneous collection of in-development and unsupported performance analysis tools for Linux ftrace and perf_events (aka the "perf" command). Both ftrace and perf are core Linux tracing tools, included in the kernel source. Your system probably has ftrace already, and perf is often just a package add (see Prerequisites). These tools are designed to be easy to install (fewest dependencies), pr
yass 2014/09/10
Linux

perf

ftrace

kernel
ãƒªãƒ³ã‚¯
perfã‚’ä½¿ã£ãŸPostgreSQLã®è§£æž(å‰ç·¨)
2. Agenda uïµâ€¯å‰ç·¨ 1.â€¯ ã¯ã˜ã‚ã« 2.â€¯ ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚°ã¨ã¯ï¼Ÿ 3.â€¯ perfã¨ã¯ 4.â€¯ perfã®ä»•çµ„ã¿ 5.â€¯ perfã¨ä»–ã®ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹è§£æžãƒ„ãƒ¼ãƒ«ã¨ã®â½æ¯”è¼ƒ 6.â€¯ perfã®å°Žâ¼Šå…¥ã¨ä½¿ã„â½…æ–¹ 7.â€¯ perfã‚’ä½¿ã£ã¦ã¿ã‚ˆã† 8.â€¯ ã¾ã¨ã‚ Copyright Â© 2012NTT DATA Corporation 2 3. Agenda uïµâ€¯å¾Œç·¨ 1.â€¯ å‰ç·¨ã®å¾©ï¥¦ç¿’ 2.â€¯ perfã‚’ä½¿ã£ãŸãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚°ã®æµï§Šã‚Œ 3.â€¯ PostgreSQLã‚’ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚°ã—ã¦ã¿ã‚ˆã† 4.â€¯ perfã®ä½¿ã„ã©ã“ã‚ 5.â€¯ ã¾ã¨ã‚ Copyright Â© 2012NTT DATA Corporation 3
yass 2014/03/20
perf

PostgreSQL
ãƒªãƒ³ã‚¯
Structured Data | Linux 6 Transparent Huge Pages and Hadoop Workloads
This past week I spent some time setting up and running various Hadoop workloads on my CDH cluster. After some Hadoop jobs had been running for several minutes, I noticed something quite alarming â€“ the system CPU percentages where extremely high. Platform Details This cluster is comprised of 2s8c16t Xeon L5630 nodes with 96 GB of RAM running CentOS Linux 6.2 with java 1.6.0_30. The details of thos
yass 2013/08/23
CDH

hadoop

hugepage

THP

perf
ãƒªãƒ³ã‚¯
Command to measure TLB misses on LINUX?
Could some one direct me to a command to measure TLB misses on LINUX, please? Is it okay to consider (or approximate) minor page faults as TLB misses?
yass 2013/04/27
perf

tlb

cache

memory

profiling
ãƒªãƒ³ã‚¯
Using Performance Counters for Linux â€“ Anton's Blog
The 2.6.31 Linux kernel will add a new performance counter subsystem called Performance Counters for Linux (or perfcounters for short). To use perfcounters, build a kernel with: CONFIG_PERF_COUNTERS=y You will need elfutils and optionally binutils (for c++ function unmangling). On debian or ubuntu: apt-get install libelf-dev binutils-dev The tools must be built 64bit on a 64bit kernel. If you have
yass 2012/12/29
perf

linux

monitoring
ãƒªãƒ³ã‚¯
https://perf.wiki.kernel.org/index.php/Tutorial
yass 2012/12/10
perf

profiling

linux
ãƒªãƒ³ã‚¯
Linux Profiling tools and techniques
Profiling is an alternative to benchmarking that is often more effective, as it gives you more fine grained measurements for the components of the system you're measuring, thus minimising external influences from consideration. It also gives the relative cost of various components, further discounting external influence. As a consequence of giving more fine grained information for a component, pro
yass 2012/12/09
perf

profiling

linux

toread
ãƒªãƒ³ã‚¯
jrfonseca - Overview
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert
yass 2012/12/09
oprofile

graph

perf

profiling
ãƒªãƒ³ã‚¯
CPU Flame Graphs
Recent posts: 29 Oct 2024 Â» AI Flame Graphs 22 Jul 2024 Â» No More Blue Fridays 24 Mar 2024 Â» Linux Crisis Tools 17 Mar 2024 Â» The Return of the Frame Pointers 10 Mar 2024 Â» eBPF Documentary 28 Apr 2023 Â» eBPF Observability Tools Are Not Security Tools 01 Mar 2023 Â» USENIX SREcon APAC 2022: Computing Performance: What's on the Horizon 17 Feb 2023 Â» USENIX SREcon APAC 2023: CFP 02 May 2022 Â» Brendan
yass 2012/12/09
linux

profiling

perf

visualization

graph
ãƒªãƒ³ã‚¯
Google Code Archive - Long-term storage for Google Code Project Hosting.
Code Archive Skip to content Google About Google Privacy Terms
yass 2012/12/09
CPU

profiling

perf
ãƒªãƒ³ã‚¯
TIP1 - Overview of C/C++ Debugging/Tracing/Profiling Tools
This document provides an overview and comparison of various tools for debugging, profiling, and tracing C/C++ programs on Linux. It discusses concepts and implementations of debugging tools like gdb and Valgrind, profiling tools like gprof and perf, and tracing tools like strace, SystemTap, and LTTng. For each tool, it provides examples of common use cases and highlights strengths and limitations
yass 2012/12/09
oprofile

perf

profiling

strace

ftrace
ãƒªãƒ³ã‚¯
ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹è§£æžã«é–¢ã™ã‚‹ã¾ã¨ã‚ - Risky Dune
ã“ã“ï¼’æ—¥ãã‚‰ã„èª¿ã¹ãŸãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹è§£æžã«é–¢ã™ã‚‹èª¿æŸ»ã‚’ã¾ã¨ã‚ã‚‹. æŽ¨æ¸¬ã‚‚å«ã¾ã‚Œã¦ã„ã‚‹ã®ã§æ³¨æ„. ã¾ãš, å½“åˆã®ç›®çš„ã¯ã€ŒBoost.Pythonã‚’ç”¨ã„ã¦ä½œã£ãŸå…±æœ‰ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã‚’Pythonã‹ã‚‰å‘¼ã³å‡ºã—ãŸéš›ã®ãƒœãƒˆãƒ«ãƒãƒƒã‚¯(hot spot)ã‚’ç™ºè¦‹ã—ãŸã„ã€ã ã£ãŸ. ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹è§£æžãƒ„ãƒ¼ãƒ«ã‚’ã„ãã¤ã‹èª¿ã¹ãŸã¨ã“ã‚, æ¬¡ã®ã‚ˆã†ãªå€™è£œãŒè€ƒãˆã‚‰ã‚ŒãŸ. oprofile google-perftools perf sprof æ±‚ã‚ã‚‹æ¡ä»¶ã¯ã€Œå…±æœ‰ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã®è§£æžãŒã§ãã‚‹ã“ã¨ã€ã¨ã€Œcall graphã‚’visualizeã™ã‚‹kcachegrindå½¢å¼ã®ãƒ‡ãƒ¼ã‚¿ã‚’å‡ºåŠ›ã§ãã‚‹ã“ã¨ã€ã®2ç‚¹ã§ã‚ã‚‹. çµå±€ã¯, google-perftoolsã‚’ä½¿ã£ã¦ã‚ã§ãŸã—ã‚ã§ãŸã—ãªã®ã ãŒ, ãã®éŽç¨‹ã§èª¿ã¹ãŸã“ã¨ã‚’ã¾ã¨ã‚ã‚‹. sprof ã“ã®è¨˜äº‹ã«ã‚ˆã‚‹ã¨gprofã¯shared libraryã®è§£æžã«ã¯ä½¿ãˆãªã„ã‚‰ã—ã„ã®ã§, sprofã®ä½¿
yass 2012/12/09
perf

oprofile

profiling
ãƒªãƒ³ã‚¯
Memory Access Patterns Are Important
In high-performance computing it is often said that the cost of a cache-miss is the largest performance penalty for an algorithm. Â For many years the increase in speed of our processors has greatly outstripped latency gains to main-memory. Â Bandwidth to main-memory has greatly increased via wider, and multi-channel, buses however the latency has not significantly reduced. Â To hide this latency our
yass 2012/12/08
cache

CPU

perf

memory

profiling

java

cache line

TLB

prefetch

hugepage
ãƒªãƒ³ã‚¯
perf statã§L1,L2(,L3)ã‚ãƒ£ãƒƒã‚·ãƒ¥ãƒŸã‚¹æ¸¬å®š - ä¿ºã¨ãŠå‰ã¨laysakura
ã“ã®è¨˜äº‹ã¯â†“ã«ç§»è»¢ã—ã¾ã—ãŸã€‚ laysakura.github.io
yass 2012/12/03
linux

perf

profiling

cpu

cache
ãƒªãƒ³ã‚¯
perf stat ã§ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã‚«ã‚¦ãƒ³ã‚¿ã®å€¤ã‚’ç›´æŽ¥æŒ‡å®šã—è¡¨ç¤º - ä¿ºã¨ãŠå‰ã¨laysakura
ã“ã®è¨˜äº‹ã¯â†“ã«ç§»è»¢ã—ã¾ã—ãŸã€‚ laysakura.github.io
yass 2012/12/03
linux

perf

profiling

cpu

cache

hardware
ãƒªãƒ³ã‚¯
perfã®ä½¿ã„ã‹ãŸ
æ¦‚è¦ Linuxã®ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹è§£æžãƒ„ãƒ¼ãƒ«ã§ã‚ã‚‹perfã®ä½¿ã„ã‹ãŸã®ç´¹ä»‹ èƒŒæ™¯ å€‹äººçš„ã«perfã‚ˆãã§ãã¦ã‚‹ã¨æ€ã†ã®ã§ç´¹ä»‹ã—ãŸã„ã¨ã„ã†ã®ã¨ã€ ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã‚«ã‚¦ãƒ³ã‚¿ã®èªã¿æ–¹ã£ã¦ã‚ã‚“ã¾ã‚ŠçŸ¥ã‚‰ã‚Œã¦ãªã„ã¿ãŸã„ãªã®ã§ã€ ãã‚Œã®è§£èª¬ã‚’æ›¸ããŸã„ã€‚ æ§‹æˆ perf ã«ã¤ã„ã¦èª¬æ˜Žã—ãŸã‚ã¨ã€ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã‚«ã‚¦ãƒ³ã‚¿ã®èªã¿ã‹ãŸã€è¦‹ã‹ãŸã€ã‚’èª¬æ˜Žã™ã‚‹ã€‚ perfã¨ã¯ä½•ã‹ Linuxã«ä»˜ã„ã¦ãã‚‹ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒ©ã€‚ man perf ã«ã‚ˆã‚‹ã¨ã€ NAME ---- perf - Performance analysis tools for Linux ã¨ã€æ›¸ã„ã¦ã‚ã‚‹ã€‚åå‰ãŒã²ã©ã„ã®ã§ãªã‚“ã¨ã‹ã—ã¦ã»ã—ã„ã€‚ perf ã®ç‰¹å¾´ å€‹äººçš„ã«ã¯ã€æ‰‹è»½ã«ä½¿ãˆã‚‹ã®ãŒç´ æ™´ã‚‰ã—ã„ã¨æ€ã†ã€‚ 2.6.31ä»¥é™ã‚«ãƒ¼ãƒãƒ«ã«æ¨™æº–ã§ä»˜ã„ã¦ã‚‹ã€‚(Ubuntuã ã¨linux-tools-common(TODO:ã‚ã¨ã§ç¢ºèª)ã§å…¥ã‚‹ã¯ãš) ç‰¹æ®Šãªè¨å®šãŒå¿…è¦ç„¡ãã€
yass 2011/10/02
linux

profiling

perf

debug
ãƒªãƒ³ã‚¯
ã¯ã¦ãªã‚°ãƒ«ãƒ¼ãƒ—ã®çµ‚äº†æ—¥ã‚’2020å¹´1æœˆ31æ—¥(é‡‘)ã«æ±ºå®šã—ã¾ã—ãŸ - ã¯ã¦ãªã®å‘ŠçŸ¥
ã¯ã¦ãªã‚°ãƒ«ãƒ¼ãƒ—ã®çµ‚äº†æ—¥ã‚’2020å¹´1æœˆ31æ—¥(é‡‘)ã«æ±ºå®šã—ã¾ã—ãŸ ä»¥ä¸‹ã®ã‚¨ãƒ³ãƒˆãƒªã®é€šã‚Šã€ä»Šå¹´æœ«ã‚’ç›®å‡¦ã«ã¯ã¦ãªã‚°ãƒ«ãƒ¼ãƒ—ã‚’çµ‚äº†äºˆå®šã§ã‚ã‚‹æ—¨ã‚’ãŠçŸ¥ã‚‰ã›ã—ã¦ãŠã‚Šã¾ã—ãŸã€‚ 2019å¹´æœ«ã‚’ç›®å‡¦ã«ã€ã¯ã¦ãªã‚°ãƒ«ãƒ¼ãƒ—ã®æä¾›ã‚’çµ‚äº†ã™ã‚‹äºˆå®šã§ã™ - ã¯ã¦ãªã‚°ãƒ«ãƒ¼ãƒ—æ—¥è¨˜ ã“ã®ãŸã³ã€æ£å¼ã«çµ‚äº†æ—¥ã‚’æ±ºå®šã„ãŸã—ã¾ã—ãŸã®ã§ã€ä»¥ä¸‹ã®é€šã‚Šã”ç¢ºèªãã ã•ã„ã€‚ çµ‚äº†æ—¥: 2020å¹´1æœˆ31æ—¥(é‡‘) ã‚¨ã‚¯ã‚¹ãƒãƒ¼ãƒˆå¸Œæœ›ç”³è«‹æœŸé™:2020å¹´1æœˆ31æ—¥(é‡‘) çµ‚äº†æ—¥ä»¥é™ã¯ã€ã¯ã¦ãªã‚°ãƒ«ãƒ¼ãƒ—ã®é–²è¦§ãŠã‚ˆã³æŠ•ç¨¿ã¯è¡Œãˆã¾ã›ã‚“ã€‚æ—¥è¨˜ã®ã‚¨ã‚¯ã‚¹ãƒãƒ¼ãƒˆãŒå¿…è¦ãªæ–¹ã¯ä»¥ä¸‹ã®è¨˜äº‹ã«ã—ãŸãŒã£ã¦æ‰‹ç¶šãã‚’ã—ã¦ãã ã•ã„ã€‚ ã¯ã¦ãªã‚°ãƒ«ãƒ¼ãƒ—ã«æŠ•ç¨¿ã•ã‚ŒãŸæ—¥è¨˜ãƒ‡ãƒ¼ã‚¿ã®ã‚¨ã‚¯ã‚¹ãƒãƒ¼ãƒˆã«ã¤ã„ã¦ - ã¯ã¦ãªã‚°ãƒ«ãƒ¼ãƒ—æ—¥è¨˜ ã”åˆ©ç”¨ã®ã¿ãªã•ã¾ã«ã¯ã”è¿·æƒ‘ã‚’ãŠã‹ã‘ã„ãŸã—ã¾ã™ãŒã€ã©ã†ãžã‚ˆã‚ã—ããŠé¡˜ã„ã„ãŸã—ã¾ã™ã€‚ 2020-06-25 è¿½è¨˜ ã¯ã¦ãªã‚°ãƒ«ãƒ¼ãƒ—æ—¥è¨˜ã®ã‚¨ã‚¯ã‚¹ãƒãƒ¼ãƒˆãƒ‡ãƒ¼ã‚¿ã¯2020å¹´2æœˆ28
yass 2011/08/21
perf

cpu

hardware

monitoring

profiling

linux
ãƒªãƒ³ã‚¯
1