Skip to content

Latest commit

 

History

History
121 lines (88 loc) · 4.38 KB

README.md

File metadata and controls

121 lines (88 loc) · 4.38 KB
Raw

profile-bee 🐝🦀🔥

Profile Bee is an eBPF based CPU profiler written in Rust for performance and efficiency.

Aya is used for building the BPF programs that is attached to perf events for sampling. The beauty of this tool is that it runs from a single binary without the need to install additional libraries such as bcctools or libbpf on target hosts.

Stacktraces are retrieved in the user space program and symbols resolution is handled.

Stacks are counted and sorted and can be output with the follow choices

  • A SVG flamegraph (generated with inferno) you can load in your browser
  • Branden Gregg's Stack Collapsed format that can be loaded up using speedscope visualizer
  • D3 flamegraph JSON and HTML output
  • Your own custom format

Note: if symbols for your C/Rust programs doesn't appear correct, you may want to build your software with debug information.

With rustc that's adding a -g flag when you compile. Another thing to consider doing is emitting frame pointer by setting RUSTFLAGS="-Cforce-frame-pointers=yes" with building (or modifying ./cargo/config) and -fno-omit-frame-pointer for gcc. With framepointers, you could get symbols while saving on the cost of dwarf parsing (using --no-dwarf)

Usage

# Profile system wide for 5s, generating a html flamegraph
profile-bee --time 5000 --html flamegraphs.html

# Profile at 9999hz for 2s, writing output to profile.svg
profile-bee --svg profile.svg --frequency 9999 --time 2000

# Realtime flamegraphs
profile-bee --time 5000 --serve --skip-idle --stream-mode 1 # Goto http://localhost:8000/ and click "realtime-updates"

# Same as above, grouped by CPU ids
profile-bee --svg profile.svg --frequency 9999 --time 2000 --group-by-cpu

# Profile at 999hz for 10s, writing output to profile.txt
profile-bee --collapse profile.txt --frequency 999 --time 10000

# Kitchen sink of all output formats
profile-bee --time 5000 --html flamegraphs.html --json profile.json --collapse profile.txt --svg profile.svg

# Profile at 99hz for 5s, writing output to screen, idle CPU cycles not counted
cargo xtask run --release -- --collapse profile.txt --frequency 99 --time 5000 --skip-idle

# Profile using kprobe over a short interval of 200ms
profile-bee --kprobe vfs_write --time 200 --svg kprobe.svg

# Profile using a tracepoint over a interval of 200ms
profile-bee --tracepoint tcp:tcp_probe --time 200 --svg tracepoint.svg

# Profile specific pid (or cpu)
profile-bee --pid <pid> ...

Features

  • Rust and C++ symbols demangling supported (via gimli)
  • Some source mapping supported (via addr2line)
  • Simple symbol lookup cache
  • SVG Flamegraph generation (via inferno)
  • BPF based stacktrace aggregation for reducing kernel <-> userspace transfers
  • Basic Kernel and tracepoint probing
  • Group by CPUs
  • Profile target PIDs, CPU id, or itself
  • Static d3 flamegraph JSON and/or HTML output
  • Real time flamegraphs served over integrated web server (using warp)

Limitations

  • Linux only
  • Interpreted / JIT stacktraces not yet supported
  • VDSO and binary offsets not calculated

TODOs

  • Optimize CPU usage
  • Check stack correctness (compare with perf, pprof etc)
  • switch over to Perf buffers
  • implement uprobing/USDT
  • pid nesting
  • Off CPU profiling
  • Optimize symbol lookup via binary search
  • Publish to crates.io
  • Measure cache hit ratio
  • Stacktrace and Hashmap clearing
  • Missing symbols

Alternatives

Development

Prerequisites

  1. Install a rust stable toolchain: rustup install stable
  2. Install a rust nightly toolchain: rustup install nightly
  3. Install bpf-linker: cargo install bpf-linker

Build eBPF

cargo xtask build-ebpf

To perform a release build you can use the --release flag. You may also change the target architecture with the --target flag

Build Userspace

cargo build

Run

cargo xtask run