1writeback(8) System Manager's Manual writeback(8)
2
3
4
6 writeback.bt - Trace file system writeback events with details. Uses
7 bpftrace/eBPF.
8
10 writeback.bt
11
13 This traces when file system dirtied pages are flushed to disk by ker‐
14 nel writeback, and prints details including when the event occurred,
15 and the duration of the event. This can be useful for correlating these
16 times with other performance problems, and if there is a match, it
17 would be a clue that the problem may be caused by writeback. How
18 quickly the kernel does writeback can be tuned: see the kernel docs,
19 eg, vm.dirty_writeback_centisecs.
20
21 This uses the tracepoint:writeback:writeback_start and trace‐
22 point:writeback:writeback_written tracepoints.
23
24 Since this uses BPF, only the root user can use this tool.
25
27 CONFIG_BPF and bpftrace.
28
30 Trace all writeback events with timestamps and latency details:
31 # writeback.bt
32
34 TIME Time that the writeback event completed, in %H:%M:%S format.
35
36 DEVICE Device name in major:minor number format.
37
38 PAGES Pages written during writeback.
39
40 REASON Reason for the writeback event. This may be "background", "vm‐
41 scan", "sync", "periodic", etc.
42
43 ms Duration of the writeback event in milliseconds.
44
46 Since writeback events are expected to be infrequent (<10/sec), the
47 overhead of this tool is expected to be negligible (near 0%).
48
50 This is from bpftrace.
51
52 https://github.com/iovisor/bpftrace
53
54 Also look in the bpftrace distribution for a companion _examples.txt
55 file containing example usage, output, and commentary for this tool.
56
58 Linux
59
61 Unstable - in development.
62
64 Brendan Gregg
65
67 biosnoop(8)
68
69
70
71USER COMMANDS 2018-09-14 writeback(8)