Blktrace是一个用户态的工具，用来收集磁盘IO信息中当IO进行到块设备层（block层，所以叫blk trace）时的详细信息（如IO请求提交，入队，合并，完成等等一些列的信息）。
块设备层处于下图（借用褚霸的图）中的 “block layer”
(1) blktrace测试的时候，会分配物理机上逻辑cpu个数个线程，并且每一个线程绑定一个逻辑cpu来收集数据
(2) blktrace在debugfs挂载的路径（默认是/sys/kernel/debug ）下每个线程产生一个文件（就有了对应的文件描述符），然后调用ioctl函数（携带文件描述符， _IOWR(0x12,115,struct blk_user_trace_setup)，& blk_user_trace_setup三个参数），产生系统调用将这些东西给内核去调用相应函数来处理，由内核经由debugfs文件系统往此文件描述符写入数据
(3) blktrace需要结合blkparse来使用，由blkparse来解析blktrace产生的特定格式的二进制数据
(4) blkparse仅打开blktrace产生的文件，从文件里面取数据做展示以及最后做per cpu的统计输出，但blkparse中展示的数据状态（如 A，U，Q，详细见下）是blkparse在t->action & 0xffff之后自己把数值转换为“A，Q，U之类的状态”来展示的。
1. yum install blktrace
2. 源码获取（你也可以从源码安装）
git clone git://git.kernel.org/pub/scm/linux/kernel/git/axboe/blktrace.git bt
cd bt
make
make install
由之前的blktrace工作原理可知，blktrace需要借助内核经由debugfs文件系统（debugfs文件系统在内存中）来输出信息
所以用blktrace工具之前需要先挂载debugfs文件系统
mount –t debugfs debugfs /sys/kernel/debug
或者在/etc/fstab中添加下面一行以便在开机启动的时候自动挂载
debug /sys/kernel/debug debugfs default 0 0
blktrace具体语法man blktrace，这里讲常用的
mkdir test #blktrace生成的数据默认会在当前目录，如之前在blktrace原理中提到，每个逻辑cpu都有一个线程，产生一个文件，故会产生cpu数目个文件
blktrace –d /dev/sda –o test1
#对 /dev/sda的trace，输出文件名为test1. Blktrace.[0-cpu数-1] （文件里面存的是二进制数据，需要blkparse来解析）
Blktrace –d /dev/sda –o - |blkparse -i –
输出到终端用“-”表示，可是都是一堆二进制东西，没法看，所以需要实时blkparse来解析
Blkparse 的“-i”后加文件名，blktrace输出为“-“代表终端（代码里面写死了，就是用这个符号来代表终端），blkparse也用“-”来代表终端解析
blkparse具体语法man blkparse，这里讲常用的
blkparse -i test1 #对test1.blktrace. [0-cpu数-1]都解析（只统计有数据的），
实时数据的解析即上blktrace的“终端输出”
终端1：
blktrace /dev/sda -o - |blkparse -i – 跑着
终端2：
dd if=/dev/zero of=/root/a1 bs=4k count=1000
终端1显示
8,0 16 3041 94.435078912 891 A W 72411584 + 8 <- (8,2) 71884224
8,0 16 3042 94.435079691 891 Q W 72411584 + 8 [flush-8:0]
8,0 16 3043 94.435080790 891 M W 72411584 + 8 [flush-8:0]
8,0 16 3044 94.435083089 891 A W 72411592 + 8 <- (8,2) 71884232
这是默认输出格式，代码里默认输出格式为，再按action输出或不输出后续信息
其中每个字母代表意思如下，数字代表占几个字符，和printf里的数字输出一样的
如
8,0 16 3042 94.435079691 891 Q W 72411584 + 8 [flush-8:0]
由于默认格式为先输出–f "%D %2c %8s %5T.%9t %5p %2a %3d "
（1）8,0 按默认输出对应%D，主从设备号
（2）16 按默认输出对应%2c，表示cpu id
（3）3042 按默认输出对应%8s，表示序列号（序列号是blkparse自己产生的一个序号，实际IO里没有这个号）
（4）94.435079691 按默认对应%5T.%9t，表示”秒.纳秒”
（5）891对应%5p,表示，进程id
（6）Q对应%2a，表示Action，Action表格如下（如Q表示IO handled by request queue code），更详细的含义见附录action表
The following table shows the various actions which may be output.
Act Description
A IO was remapped to a different device
B IO bounced
C IO completion
D IO issued to driver
F IO front merged with request on queue
G Get request
I IO inserted onto request queue
M IO back merged with request on queue
P Plug request
Q IO handled by request queue code
S Sleep request
T Unplug due to timeout
U Unplug request
X Split
（7）W 对应%3d，表示RWBS域（W表示写操作），各字母含义如下
至少包含“RWD“（ R 读，W写，D块被忽略）中的1个字符
还可以附加“BS“（B barrier，S同步）
switch (act[0]) {
case 'R': /* Requeue */
case 'C': /* Complete */
if (t->action & BLK_TC_ACT(BLK_TC_PC)) {
char *p = dump_pdu(pdu_buf, pdu_len);
if (p)
fprintf(ofp, "(%s) ", p);
fprintf(ofp, "[%d]n", t->error);
} else {
if (elapsed != -1ULL) {
if (t_sec(t))
fprintf(ofp, "%llu + %u (%8llu) [%d]n",
(unsigned long long) t->sector,
t_sec(t), elapsed, t->error);
else
fprintf(ofp, "%llu (%8llu) [%d]n",
(unsigned long long) t->sector,
elapsed, t->error);
} else {
if (t_sec(t))
fprintf(ofp, "%llu + %u [%d]n",
(unsigned long long) t->sector,
t_sec(t), t->error);
else
fprintf(ofp, "%llu [%d]n",
(unsigned long long) t->sector,
t->error);
}
}
break;
case 'D': /* Issue */
case 'I': /* Insert */
case 'Q': /* Queue */
case 'B': /* Bounce */
if (t->action & BLK_TC_ACT(BLK_TC_PC)) {
char *p;
fprintf(ofp, "%u ", t->bytes);
p = dump_pdu(pdu_buf, pdu_len);
if (p)
fprintf(ofp, "(%s) ", p);
fprintf(ofp, "[%s]n", name);
} else {
if (elapsed != -1ULL) {
if (t_sec(t))
fprintf(ofp, "%llu + %u (%8llu) [%s]n",
(unsigned long long) t->sector,
t_sec(t), elapsed, name);
else
fprintf(ofp, "(%8llu) [%s]n", elapsed,
name);
} else {
if (t_sec(t))
fprintf(ofp, "%llu + %u [%s]n",
(unsigned long long) t->sector,
t_sec(t), name);
else
fprintf(ofp, "[%s]n", name);
}
}
break;
case 'M': /* Back merge */
case 'F': /* Front merge */
case 'G': /* Get request */
case 'S': /* Sleep request */
if (t_sec(t))
fprintf(ofp, "%llu + %u [%s]n",
(unsigned long long) t->sector, t_sec(t), name);
else
fprintf(ofp, "[%s]n", name);
break;
case 'P': /* Plug */
fprintf(ofp, "[%s]n", name);
break;
case 'U': /* Unplug IO */
case 'T': /* Unplug timer */
fprintf(ofp, "[%s] %un", name, get_pdu_int(t));
break;
case 'A': /* remap */
get_pdu_remap(t, &r);
fprintf(ofp, "%llu + %u <- (%d,%d) %llun",
(unsigned long long) t->sector, t_sec(t),
MAJOR(r.device_from), MINOR(r.device_from),
(unsigned long long) r.sector_from);
break;
case 'X': /* Split */
fprintf(ofp, "%llu / %u [%s]n", (unsigned long long) t->sector,
get_pdu_int(t), name);
break;
case 'm': /* Message */
fprintf(ofp, "%*sn", pdu_len, pdu_buf);
break;
default:
fprintf(stderr, "Unknown action %cn", act[0]);
break;
}
所以
8,0 16 3042 94.435079691 891 Q W 72411584 + 8 [flush-8:0]
中的act[0]=’Q’,后面的72411584是（8，0即sda）相对8:0的扇区起始号，+8，为后面连续的8个扇区（默认一个扇区512byte，所以8个扇区就是4K），后面的[flush-8:0]是程序的名字。
8,0 16 3041 94.435078912 891 A W 72411584 + 8 <- (8,2) 71884224
Action[0]=’A’, 72411584是相对8:0（即sda）的起始扇区号，（8,2）是相对/dev/sda2分区的扇区号为71884224，(由于/dev/sda2分区时sda磁盘上面的一个分区，故sda2上面的起始位置要先映射到sda磁盘上面去)
由于扇区号在磁盘上面是连续的，磁盘又被格式化成很多块，一个块里包含多个扇区，所以，扇区号/块大小=块号，
根据块号你就可以找到对应的inode，
debugfs -R 'icheck 块号' 具体磁盘或分区
如你的扇区号是相对sda2上面算出来的块号，那debugfs –R ‘icheck 块号’ /dev/sda2就可以找到对应的inode
根据inode你就可以找到对应的文件是什么了
find / -inum your_inode
有一个例子见淘宝牛人写的一篇链接地址
C – complete A previously issued request has been completed. The output
will detail the sector and size of that request, as well as the success or
failure of it.
D – issued A request that previously resided on the block layer queue or in
the io scheduler has been sent to the driver.
I – inserted A request is being sent to the io scheduler for addition to the
internal queue and later service by the driver. The request is fully formed
at this time.
Q – queued This notes intent to queue io at the given location. No real requests
exists yet.
B – bounced The data pages attached to this bio are not reachable by the
hardware and must be bounced to a lower memory location. This causes
a big slowdown in io performance, since the data must be copied to/from
kernel buffers. Usually this can be fixed with using better hardware -
either a better io controller, or a platform with an IOMMU.
m – message Text message generated via kernel call to blk add trace msg.
M – back merge A previously inserted request exists that ends on the boundary
of where this io begins, so the io scheduler can merge them together.
F – front merge Same as the back merge, except this io ends where a previously
inserted requests starts.
G – get request To send any type of request to a block device, a struct request
container must be allocated first.
S – sleep No available request structures were available, so the issuer has to
wait for one to be freed.
P – plug When io is queued to a previously empty block device queue, Linux
will plug the queue in anticipation of future ios being added before this
data is needed.
U – unplug Some request data already queued in the device, start sending
requests to the driver. This may happen automatically if a timeout period
has passed (see next entry) or if a number of requests have been added to
the queue.
T – unplug due to timer If nobody requests the io that was queued after
plugging the queue, Linux will automatically unplug it after a defined
period has passed.
X – split On raid or device mapper setups, an incoming io may straddle a
device or internal zone and needs to be chopped up into smaller pieces
for service. This may indicate a performance problem due to a bad setup
of that raid/dm device, but may also just be part of normal boundary
conditions. dm is notably bad at this and will clone lots of io.
A – remap For stacked devices, incoming io is remapped to device below it in
the io stack. The remap action details what exactly is being remapped to
what.
外带一张图，可能看得更清楚
