bcc-pythoncalls(8)

1ucalls(8)                   System Manager's Manual                  ucalls(8)
2
3
4

NAME

6       ucalls,  javacalls, pythoncalls, rubycalls, phpcalls - Summarize method
7       calls from high-level languages and Linux syscalls.
8

SYNOPSIS

10       javacalls [-h] [-T TOP] [-L] [-S] [-v] [-m] pid [interval]
11       pythoncalls [-h] [-T TOP] [-L] [-S] [-v] [-m] pid [interval]
12       rubycalls [-h] [-T TOP] [-L] [-S] [-v] [-m] pid [interval]
13       phpcalls [-h] [-T TOP] [-L] [-S] [-v] [-m] pid [interval]
14       ucalls [-l {java,python,ruby,php}] [-h] [-T TOP] [-L]  [-S]  [-v]  [-m]
15       pid [interval]
16

DESCRIPTION

18       This  tool  summarizes  method  calls from high-level languages such as
19       Python, Java, Ruby, and PHP. It can  also  trace  Linux  system  calls.
20       Whenever a method is invoked, ucalls records the call count and option‐
21       ally the method's execution time (latency) and displays a summary.
22
23       This uses in-kernel eBPF maps to store per process summaries for  effi‐
24       ciency.
25
26       This  tool relies on USDT probes embedded in many high-level languages,
27       such as Java, Python, Ruby, and PHP. It requires a runtime instrumented
28       with  these  probes,  which in some cases requires building from source
29       with a  USDT-specific  flag,  such  as  "--enable-dtrace"  or  "--with-
30       dtrace". For Java, method probes are not enabled by default, and can be
31       turned on by running the  Java  process  with  the  "-XX:+ExtendedDTra‐
32       ceProbes"   flag.   For   PHP   processes,   the  environment  variable
33       USE_ZEND_DTRACE must be set to 1.
34
35       Since this uses BPF, only the root user can use this tool.
36

REQUIREMENTS

38       CONFIG_BPF and bcc.
39

OPTIONS

41       -l {java,python,ruby,php}
42              The language to trace. If not provided, only syscalls are traced
43              (when the -S option is used).
44
45       -T TOP Print only the top methods by frequency or latency.
46
47       -L     Collect method invocation latency (duration).
48
49       -S     Collect Linux syscalls frequency and timing.
50
51       -v     Print the resulting BPF program, for debugging purposes.
52
53       -m     Print times in milliseconds (the default is microseconds).
54
55       pid    The process id to trace.
56
57       interval
58              Print  summary  after  this  number of seconds and then exit. By
59              default, wait for Ctrl+C to terminate.
60

EXAMPLES

62       Trace the top 10 Ruby method calls:
63              # ucalls -T 10 -l ruby 1344
64
65       Trace Python method calls and Linux syscalls including latency in  mil‐
66       liseconds:
67              # ucalls -l python -mL 2020
68
69       Trace only syscalls and print a summary after 10 seconds:
70              # ucalls -S 788 10
71

OVERHEAD

73       Tracing individual method calls will produce a considerable overhead in
74       all high-level languages. For languages with just-in-time  compilation,
75       such  as  Java,  the  overhead can be more considerable than for inter‐
76       preted languages.  On the other hand, syscall tracing will typically be
77       tolerable  for  most processes, unless they have a very unusual rate of
78       system calls.
79

SOURCE

81       This is from bcc.
82
83              https://github.com/iovisor/bcc
84
85       Also look in the bcc distribution for  a  companion  _example.txt  file
86       containing example usage, output, and commentary for this tool.
87

OS

89       Linux
90

STABILITY

92       Unstable - in development.
93

AUTHOR

95       Sasha Goldshtein
96