zfsslower

README.md

@@ -106,6 +106,7 @@ Tools:
 - tools/[wakeuptime](tools/wakeuptime.py): Summarize sleep to wakeup time by waker kernel stack. [Examples](tools/wakeuptime_example.txt).
 - tools/[xfsdist](tools/xfsdist.py): Summarize XFS operation latency. [Examples](tools/xfsdist_example.txt).
 - tools/[xfsslower](tools/xfsslower.py): Trace slow XFS operations. [Examples](tools/xfsslower_example.txt).
+- tools/[zfsslower](tools/zfsslower.py): Trace slow ZFS operations. [Examples](tools/zfsslower_example.txt).
 
 ### Networking
 

man/man8/zfsslower.8

+.TH zfsslower 8  "2016-02-11" "USER COMMANDS"
+.SH NAME
+zfsslower \- Trace slow zfs file operations, with per-event details.
+.SH SYNOPSIS
+.B zfsslower [\-h] [\-j] [\-p PID] [min_ms]
+.SH DESCRIPTION
+This tool traces common ZFS file operations: reads, writes, opens, and
+syncs. It measures the time spent in these operations, and prints details
+for each that exceeded a threshold.
+
+WARNING: See the OVERHEAD section.
+
+By default, a minimum millisecond threshold of 10 is used. If a threshold of 0
+is used, all events are printed (warning: verbose).
+
+This uses kernel dynamic tracing of the ZPL interface (ZFS POSIX
+Layer), and will need updates to match any changes to this interface.
+.TP
+This is intended to work with the ZFS on Linux project:
+http://zfsonlinux.org
+.PP
+Since this uses BPF, only the root user can use this tool.
+.SH REQUIREMENTS
+CONFIG_BPF and bcc.
+.SH OPTIONS
+\-p PID
+Trace this PID only.
+.TP
+min_ms
+Minimum I/O latency (duration) to trace, in milliseconds. Default is 10 ms.
+.SH EXAMPLES
+.TP
+Trace synchronous file reads and writes slower than 10 ms:
+#
+.B zfsslower
+.TP
+Trace slower than 1 ms:
+#
+.B zfsslower 1
+.TP
+Trace slower than 1 ms, and output just the fields in parsable format (csv):
+#
+.B zfsslower \-j 1
+.TP
+Trace all file reads and writes (warning: the output will be verbose):
+#
+.B zfsslower 0
+.TP
+Trace slower than 1 ms, for PID 181 only:
+#
+.B zfsslower \-p 181 1
+.SH FIELDS
+.TP
+TIME(s)
+Time of I/O completion since the first I/O seen, in seconds.
+.TP
+COMM
+Process name.
+.TP
+PID
+Process ID.
+.TP
+T
+Type of operation. R == read, W == write, O == open, S == fsync.
+.TP
+OFF_KB
+File offset for the I/O, in Kbytes.
+.TP
+BYTES
+Size of I/O, in bytes.
+.TP
+LAT(ms)
+Latency (duration) of I/O, measured from when it was issued by VFS to the
+filesystem, to when it completed. This time is inclusive of block device I/O,
+file system CPU cycles, file system locks, run queue latency, etc. It's a more
+accurate measure of the latency suffered by applications performing file
+system I/O, than to measure this down at the block device interface.
+.TP
+FILENAME
+A cached kernel file name (comes from dentry->d_iname).
+.TP
+ENDTIME_us
+Completion timestamp, microseconds (\-j only).
+.TP
+OFFSET_b
+File offset, bytes (\-j only).
+.TP
+LATENCY_us
+Latency (duration) of the I/O, in microseconds (\-j only).
+.SH OVERHEAD
+This adds low-overhead instrumentation to these ZFS operations,
+including reads and writes from the file system cache. Such reads and writes
+can be very frequent (depending on the workload; eg, 1M/sec), at which
+point the overhead of this tool (even if it prints no "slower" events) can
+begin to become significant. Measure and quantify before use. If this
+continues to be a problem, consider switching to a tool that prints in-kernel
+summaries only.
+.PP
+Note that the overhead of this tool should be less than fileslower(8), as
+this tool targets zfs functions only, and not all file read/write paths
+(which can include socket I/O).
+.SH SOURCE
+This is from bcc.
+.IP
+https://github.com/iovisor/bcc
+.PP
+Also look in the bcc distribution for a companion _examples.txt file containing
+example usage, output, and commentary for this tool.
+.SH OS
+Linux
+.SH STABILITY
+Unstable - in development.
+.SH AUTHOR
+Brendan Gregg
+.SH SEE ALSO
+biosnoop(8), funccount(8), fileslower(8)

tools/zfsslower.py

+#!/usr/bin/python
+# @lint-avoid-python-3-compatibility-imports
+#
+# zfsslower  Trace slow ZFS operations.
+#            For Linux, uses BCC, eBPF.
+#
+# USAGE: zfsslower [-h] [-j] [-p PID] [min_ms]
+#
+# This script traces common ZFS file operations: reads, writes, opens, and
+# syncs. It measures the time spent in these operations, and prints details
+# for each that exceeded a threshold.
+#
+# WARNING: This adds low-overhead instrumentation to these ZFS operations,
+# including reads and writes from the file system cache. Such reads and writes
+# can be very frequent (depending on the workload; eg, 1M/sec), at which
+# point the overhead of this tool (even if it prints no "slower" events) can
+# begin to become significant.
+#
+# This works by using kernel dynamic tracing of the ZPL interface, and will
+# need updates to match any changes to this interface.
+#
+# By default, a minimum millisecond threshold of 10 is used.
+#
+# Copyright 2016 Netflix, Inc.
+# Licensed under the Apache License, Version 2.0 (the "License")
+#
+# 14-Feb-2016   Brendan Gregg   Created this.
+
+from __future__ import print_function
+from bcc import BPF
+import argparse
+from time import strftime
+import ctypes as ct
+
+# arguments
+examples = """examples:
+    ./zfsslower             # trace operations slower than 10 ms (default)
+    ./zfsslower 1           # trace operations slower than 1 ms
+    ./zfsslower -j 1        # ... 1 ms, parsable output (csv)
+    ./zfsslower 0           # trace all operations (warning: verbose)
+    ./zfsslower -p 185      # trace PID 185 only
+"""
+parser = argparse.ArgumentParser(
+    description="Trace common ZFS file operations slower than a threshold",
+    formatter_class=argparse.RawDescriptionHelpFormatter,
+    epilog=examples)
+parser.add_argument("-j", "--csv", action="store_true",
+    help="just print fields: comma-separated values")
+parser.add_argument("-p", "--pid",
+    help="trace this PID only")
+parser.add_argument("min_ms", nargs="?", default='10',
+    help="minimum I/O duration to trace, in ms (default 10)")
+args = parser.parse_args()
+min_ms = int(args.min_ms)
+pid = args.pid
+csv = args.csv
+debug = 0
+
+# define BPF program
+bpf_text = """
+#include <uapi/linux/ptrace.h>
+#include <linux/fs.h>
+#include <linux/sched.h>
+#include <linux/dcache.h>
+
+// XXX: switch these to char's when supported
+#define TRACE_READ      0
+#define TRACE_WRITE     1
+#define TRACE_OPEN      2
+#define TRACE_FSYNC     3
+
+struct val_t {
+    u64 ts;
+    u64 offset;
+    struct file *fp;
+};
+
+struct data_t {
+    // XXX: switch some to u32's when supported
+    u64 ts_us;
+    u64 type;
+    u64 size;
+    u64 offset;
+    u64 delta_us;
+    u64 pid;
+    char task[TASK_COMM_LEN];
+    char file[DNAME_INLINE_LEN];
+};
+
+BPF_HASH(entryinfo, pid_t, struct val_t);
+BPF_PERF_OUTPUT(events);
+
+//
+// Store timestamp and size on entry
+//
+
+// zpl_read(), zpl_write():
+int trace_rw_entry(struct pt_regs *ctx, struct file *filp, char __user *buf,
+    size_t len, loff_t *ppos)
+{
+    u32 pid;
+    pid = bpf_get_current_pid_tgid();
+    if (FILTER_PID)
+        return 0;
+
+    // store filep and timestamp by pid
+    struct val_t val = {};
+    val.ts = bpf_ktime_get_ns();
+    val.fp = filp;
+    val.offset = *ppos;
+    if (val.fp)
+        entryinfo.update(&pid, &val);
+
+    return 0;
+}
+
+// zpl_open():
+int trace_open_entry(struct pt_regs *ctx, struct inode *inode,
+    struct file *filp)
+{
+    u32 pid;
+    pid = bpf_get_current_pid_tgid();
+    if (FILTER_PID)
+        return 0;
+
+    // store filep and timestamp by pid
+    struct val_t val = {};
+    val.ts = bpf_ktime_get_ns();
+    val.fp = filp;
+    val.offset = 0;
+    if (val.fp)
+        entryinfo.update(&pid, &val);
+
+    return 0;
+}
+
+// zpl_fsync():
+int trace_fsync_entry(struct pt_regs *ctx, struct file *filp)
+{
+    u32 pid;
+    pid = bpf_get_current_pid_tgid();
+    if (FILTER_PID)
+        return 0;
+
+    // store filp and timestamp by pid
+    struct val_t val = {};
+    val.ts = bpf_ktime_get_ns();
+    val.fp = filp;
+    val.offset = 0;
+    if (val.fp)
+        entryinfo.update(&pid, &val);
+
+    return 0;
+}
+
+//
+// Output
+//
+
+static int trace_return(struct pt_regs *ctx, int type)
+{
+    struct val_t *valp;
+    u32 pid = bpf_get_current_pid_tgid();
+
+    valp = entryinfo.lookup(&pid);
+    if (valp == 0) {
+        // missed tracing issue or filtered
+        return 0;
+    }
+
+    // calculate delta
+    u64 ts = bpf_ktime_get_ns();
+    u64 delta_us = (ts - valp->ts) / 1000;
+    entryinfo.delete(&pid);
+    if (FILTER_US)
+        return 0;
+
+    // workaround (rewriter should handle file to d_iname in one step):
+    struct dentry *de = NULL;
+    bpf_probe_read(&de, sizeof(de), &valp->fp->f_path.dentry);
+
+    // populate output struct
+    u32 size = ctx->ax;
+    struct data_t data = {.type = type, .size = size, .delta_us = delta_us,
+        .pid = pid};
+    data.ts_us = ts / 1000;
+    data.offset = valp->offset;
+    bpf_probe_read(&data.file, sizeof(data.file), de->d_iname);
+    bpf_get_current_comm(&data.task, sizeof(data.task));
+
+    events.perf_submit(ctx, &data, sizeof(data));
+
+    return 0;
+}
+
+int trace_read_return(struct pt_regs *ctx)
+{
+    return trace_return(ctx, TRACE_READ);
+}
+
+int trace_write_return(struct pt_regs *ctx)
+{
+    return trace_return(ctx, TRACE_WRITE);
+}
+
+int trace_open_return(struct pt_regs *ctx)
+{
+    return trace_return(ctx, TRACE_OPEN);
+}
+
+int trace_fsync_return(struct pt_regs *ctx)
+{
+    return trace_return(ctx, TRACE_FSYNC);
+}
+
+"""
+if min_ms == 0:
+    bpf_text = bpf_text.replace('FILTER_US', '0')
+else:
+    bpf_text = bpf_text.replace('FILTER_US',
+        'delta_us <= %s' % str(min_ms * 1000))
+if args.pid:
+    bpf_text = bpf_text.replace('FILTER_PID', 'pid != %s' % pid)
+else:
+    bpf_text = bpf_text.replace('FILTER_PID', '0')
+if debug:
+    print(bpf_text)
+
+# kernel->user event data: struct data_t
+DNAME_INLINE_LEN = 32   # linux/dcache.h
+TASK_COMM_LEN = 16      # linux/sched.h
+class Data(ct.Structure):
+    _fields_ = [
+        ("ts_us", ct.c_ulonglong),
+        ("type", ct.c_ulonglong),
+        ("size", ct.c_ulonglong),
+        ("offset", ct.c_ulonglong),
+        ("delta_us", ct.c_ulonglong),
+        ("pid", ct.c_ulonglong),
+        ("task", ct.c_char * TASK_COMM_LEN),
+        ("file", ct.c_char * DNAME_INLINE_LEN)
+    ]
+
+# process event
+def print_event(cpu, data, size):
+    event = ct.cast(data, ct.POINTER(Data)).contents
+
+    type = 'R'
+    if event.type == 1:
+        type = 'W'
+    elif event.type == 2:
+        type = 'O'
+    elif event.type == 3:
+        type = 'S'
+
+    if (csv):
+        print("%d,%s,%d,%s,%d,%d,%d,%s" % (
+            event.ts_us, event.task, event.pid, type, event.size,
+            event.offset, event.delta_us, event.file))
+        return
+    print("%-8s %-14.14s %-6s %1s %-7s %-8d %7.2f %s" % (strftime("%H:%M:%S"),
+        event.task, event.pid, type, event.size, event.offset / 1024,
+        float(event.delta_us) / 1000, event.file))
+
+# initialize BPF
+b = BPF(text=bpf_text)
+
+# common file functions
+b.attach_kprobe(event="zpl_read", fn_name="trace_rw_entry")
+b.attach_kprobe(event="zpl_write", fn_name="trace_rw_entry")
+b.attach_kprobe(event="zpl_open", fn_name="trace_open_entry")
+b.attach_kprobe(event="zpl_fsync", fn_name="trace_fsync_entry")
+b.attach_kretprobe(event="zpl_read", fn_name="trace_read_return")
+b.attach_kretprobe(event="zpl_write", fn_name="trace_write_return")
+b.attach_kretprobe(event="zpl_open", fn_name="trace_open_return")
+b.attach_kretprobe(event="zpl_fsync", fn_name="trace_fsync_return")
+
+# header
+if (csv):
+    print("ENDTIME_us,TASK,PID,TYPE,BYTES,OFFSET_b,LATENCY_us,FILE")
+else:
+    if min_ms == 0:
+        print("Tracing ZFS operations")
+    else:
+        print("Tracing ZFS operations slower than %d ms" % min_ms)
+    print("%-8s %-14s %-6s %1s %-7s %-8s %7s %s" % ("TIME", "COMM", "PID", "T",
+        "BYTES", "OFF_KB", "LAT(ms)", "FILENAME"))
+
+# read events
+b["events"].open_perf_buffer(print_event)
+while 1:
+    b.kprobe_poll()

tools/zfsslower_example.txt

+Demonstrations of zfsslower, the Linux eBPF/bcc version.
+
+
+zfsslower shows ZFS reads, writes, opens, and fsyncs, slower than a threshold.
+It has been written to work on ZFS on Linux (http://zfsonlinux.org). For
+example:
+
+# ./zfsslower 
+Tracing ZFS operations slower than 10 ms
+TIME     COMM           PID    T BYTES   OFF_KB   LAT(ms) FILENAME
+06:31:28 dd             25570  W 131072  38784     303.92 data1
+06:31:34 dd             25686  W 131072  38784     388.28 data1
+06:31:35 dd             25686  W 131072  78720     519.66 data1
+06:31:35 dd             25686  W 131072  116992    405.94 data1
+06:31:35 dd             25686  W 131072  153600    433.52 data1
+06:31:36 dd             25686  W 131072  188672    314.37 data1
+06:31:36 dd             25686  W 131072  222336    372.33 data1
+06:31:36 dd             25686  W 131072  254592    309.59 data1
+06:31:37 dd             25686  W 131072  285440    304.52 data1
+06:31:37 dd             25686  W 131072  315008    236.45 data1
+06:31:37 dd             25686  W 131072  343424    193.54 data1
+06:31:38 dd             25686  W 131072  370560    286.07 data1
+06:31:38 dd             25686  W 131072  396672    251.92 data1
+[...]
+
+This shows writes to a "data1" file, each taking well over the 10 ms threshold.
+the slowest, on the 3rd line of output, reached 519.66 ms for a 128 Kbyte
+write by the "dd" command.
+
+This "latency" is measured from when the operation was issued from the VFS
+interface to the file system (via the ZFS POSIX layer), to when it completed.
+This spans everything: block device I/O (disk I/O), file system CPU cycles,
+file system locks, run queue latency, etc. This is a better measure of the
+latency suffered by applications reading from the file system than measuring
+this down at the block device interface.
+
+Note that this only traces the common file system operations previously
+listed: other file system operations (eg, inode operations including
+getattr()) are not traced.
+
+
+A threshold of 0 will trace all operations. Warning: the output will be
+verbose, as it will include all file system cache hits.
+
+# ./zfsslower 0
+Tracing ZFS operations
+TIME     COMM           PID    T BYTES   OFF_KB   LAT(ms) FILENAME
+06:36:07 dd             32242  O 0       0           0.01 data1
+06:36:07 dd             32242  W 131072  0           0.25 data1
+06:36:07 dd             32242  W 131072  128         0.03 data1
+06:36:07 dd             32242  W 131072  256         0.04 data1
+06:36:07 dd             32242  W 131072  384         0.04 data1
+06:36:07 dd             32242  W 131072  512         0.04 data1
+06:36:07 dd             32242  W 131072  640         0.03 data1
+06:36:07 dd             32242  W 131072  768         0.03 data1
+06:36:07 dd             32242  W 131072  896         0.04 data1
+06:36:07 dd             32242  W 131072  1024        0.28 data1
+06:36:07 dd             32242  W 131072  1152        0.04 data1
+06:36:07 dd             32242  W 131072  1280        0.03 data1
+[...]
+06:36:07 dd             32242  W 131072  13824       0.04 data1
+06:36:07 dd             32242  W 131072  13952       0.04 data1
+06:36:07 dd             32242  W 131072  14080       0.04 data1
+06:36:07 dd             32242  W 131072  14208     398.92 data1
+06:36:07 dd             32242  W 131072  14336       0.04 data1
+06:36:07 dd             32242  W 131072  14464       0.04 data1
+06:36:07 dd             32242  W 131072  15104       0.03 data1
+[...]
+
+The output now includes the open operation for this file ("O"), and then the
+writes. Most of the writes are very fast, with only an occasional outlier that
+is in the hundreds of milliseconds.
+
+Fortunately this is not a real world environment: I setup a zpool on top of a
+XFS file system for testing purposes. More debugging using other tools will
+explain these outliers: possibly XFS flushing.
+
+
+Here's a random read workload, and showing operations slower than 1 ms:
+
+# ./zfsslower 1
+Tracing ZFS operations slower than 1 ms
+TIME     COMM           PID    T BYTES   OFF_KB   LAT(ms) FILENAME
+06:47:30 randread.pl    15431  R 8192    97840       1.03 data1
+06:47:30 randread.pl    15431  R 8192    416744      1.12 data1
+06:47:31 randread.pl    15431  R 8192    228856      1.96 data1
+06:47:31 randread.pl    15431  R 8192    452248      1.02 data1
+06:47:31 randread.pl    15431  R 8192    315288      5.90 data1
+06:47:31 randread.pl    15431  R 8192    752696      1.20 data1
+06:47:31 randread.pl    15431  R 8192    481832      1.39 data1
+06:47:31 randread.pl    15431  R 8192    673752      1.39 data1
+06:47:31 randread.pl    15431  R 8192    691736      1.01 data1
+06:47:31 randread.pl    15431  R 8192    694776      1.78 data1
+06:47:31 randread.pl    15431  R 8192    403328      3.75 data1
+06:47:31 randread.pl    15431  R 8192    567688      1.08 data1
+06:47:31 randread.pl    15431  R 8192    694280      1.31 data1
+06:47:31 randread.pl    15431  R 8192    669280      1.06 data1
+06:47:31 randread.pl    15431  R 8192    426608      1.56 data1
+06:47:31 randread.pl    15431  R 8192    42512       1.01 data1
+06:47:31 randread.pl    15431  R 8192    22944       1.33 data1
+06:47:31 randread.pl    15431  R 8192    427432      1.48 data1
+06:47:31 randread.pl    15431  R 8192    261320      1.28 data1
+06:47:31 randread.pl    15431  R 8192    132248      1.23 data1
+06:47:31 randread.pl    15431  R 8192    96936       1.04 data1
+06:47:31 randread.pl    15431  R 8192    482800      2.63 data1
+[...]
+
+
+A -j option will print just the fields (parsable output, csv):
+
+# ./zfsslower -j 1
+ENDTIME_us,TASK,PID,TYPE,BYTES,OFFSET_b,LATENCY_us,FILE
+252305490911,randread.pl,17922,R,8192,163446784,1156,data1
+252305493852,randread.pl,17922,R,8192,321437696,1129,data1
+252305498839,randread.pl,17922,R,8192,475152384,1154,data1
+252305505515,randread.pl,17922,R,8192,49094656,1082,data1
+252305506774,randread.pl,17922,R,8192,470401024,1245,data1
+252305509265,randread.pl,17922,R,8192,553246720,2412,data1
+252305512365,randread.pl,17922,R,8192,20963328,1093,data1
+252305513755,randread.pl,17922,R,8192,304111616,1350,data1
+252305583330,randread.pl,17922,R,8192,166174720,1154,data1
+252305593913,randread.pl,17922,R,8192,175079424,1241,data1
+252305602833,randread.pl,17922,R,8192,305340416,3307,data1
+252305608663,randread.pl,17922,R,8192,655958016,2704,data1
+252305611212,randread.pl,17922,R,8192,40951808,1033,data1
+252305614609,randread.pl,17922,R,8192,318922752,2687,data1
+252305623800,randread.pl,17922,R,8192,246734848,2983,data1
+252305711125,randread.pl,17922,R,8192,581795840,1091,data1
+252305728694,randread.pl,17922,R,8192,710483968,1034,data1
+252305762046,randread.pl,17922,R,8192,329367552,1405,data1
+252305798215,randread.pl,17922,R,8192,44482560,1030,data1
+252305806748,randread.pl,17922,R,8192,660602880,1069,data1
+252305826360,randread.pl,17922,R,8192,616144896,2327,data1
+[...]
+
+
+USAGE message:
+
+# ./zfsslower -h
+usage: zfsslower [-h] [-j] [-p PID] [min_ms]
+
+Trace common ZFS file operations slower than a threshold
+
+positional arguments:
+  min_ms             minimum I/O duration to trace, in ms (default 10)
+
+optional arguments:
+  -h, --help         show this help message and exit
+  -j, --csv          just print fields: comma-separated values
+  -p PID, --pid PID  trace this PID only
+
+examples:
+    ./zfsslower             # trace operations slower than 10 ms (default)
+    ./zfsslower 1           # trace operations slower than 1 ms
+    ./zfsslower -j 1        # ... 1 ms, parsable output (csv)
+    ./zfsslower 0           # trace all operations (warning: verbose)
+    ./zfsslower -p 185      # trace PID 185 only