DOKK / manpages / debian 12 / bpfcc-tools / nodestat-bpfcc.8.en
ustat(8) System Manager's Manual ustat(8)

ustat, javastat, nodestat, perlstat, phpstat, pythonstat, rubystat, tclstat - Activity stats from high-level languages.

javastat [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]
nodestat [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]
perlstat [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]
phpstat [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]
pythonstat [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]
rubystat [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]
tclstat [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]
ustat [-l {java,node,perl,php,python,ruby,tcl}] [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]

This is "top" for high-level language events, such as garbage collections, exceptions, thread creations, object allocations, method calls, and more. The events are aggregated for each process and printed in a top-like table, which can be sorted by various fields. Not all language runtimes provide the same set of details.

This uses in-kernel eBPF maps to store per process summaries for efficiency.

This tool relies on USDT probes embedded in many high-level languages, such as Java, Node, Perl, PHP, Python, Ruby, and Tcl. It requires a runtime instrumented with these probes, which in some cases requires building from source with a USDT-specific flag, such as "--enable-dtrace" or "--with-dtrace". For Java, some probes are not enabled by default, and can be turned on by running the Java process with the "-XX:+ExtendedDTraceProbes" flag. For PHP processes, the environment variable USE_ZEND_DTRACE must be set to 1.

Newly-created processes will only be traced at the next interval. If you run this tool with a short interval (say, 1-5 seconds), this should be virtually unnoticeable. For longer intervals, you might miss processes that were started and terminated during the interval window.

Since this uses BPF, only the root user can use this tool.

CONFIG_BPF and bcc.

The language to trace. By default, all languages are traced.
Do not clear the screen between updates.
Sort the output by the specified field.
Do not print more than this number of rows.
Print the resulting BPF program, for debugging purposes.
Interval between updates, seconds.
Number of interval summaries.

# ustat
# ustat -Cr 8
5 second summaries, 10 times only:
# ustat 5 10

The contents of /proc/loadavg
Process ID.
Process command line (often the second and following arguments will give you a hint as to which application is being run.
Count of method invocations during interval.
Count of garbage collections during interval.
Count of objects allocated during interval.
Count of classes loaded during interval.
Count of exceptions thrown during interval.
Count of threads created during interval.

When using this tool with high-frequency events, such as method calls, a very significant slow-down can be expected. However, many of the high-level languages covered by this tool already have a fairly high per-method invocation cost, especially when running in interpreted mode. For the lower-frequency events, such as garbage collections or thread creations, the overhead should not be significant. Specifically, when probing Java processes and not using the "-XX:+ExtendedDTraceProbes" flag, the most expensive probes are not emitted, and the overhead should be acceptable.

This is from bcc.

https://github.com/iovisor/bcc

Also look in the bcc distribution for a companion _example.txt file containing example usage, output, and commentary for this tool.

Linux

Unstable - in development.

Sasha Goldshtein

trace(8), argdist(8), tplist(8)

2018-10-09 USER COMMANDS