什麼是iowait?
顧名思義,就是系統因爲io導致的進程wait。再深一點講就是:這時候系統在做io,導致沒有進程在幹活,cpu在執行idle進程空轉,所以說iowait的產生要滿足兩個條件,一是進程在等io,二是等io時沒有進程可運行。
Iowait是如何計算的?
先說說用戶如何看到iowait吧
我們通常用vmstat就能看到iowat,圖中的wa就是(標紅)
這個數據是vmstat經過計算文件/proc/stat中的數據獲得,所以說大家看到的是能夠大概反應一個系統iowait水平的數據表象。關於/proc/stat中的數據都代表了什麼意思,大家自己google吧,不再贅述。
for_each_possible_cpu(i) { intj; user = cputime64_add(user, kstat_cpu(i).cpustat.user); nice = cputime64_add(nice, kstat_cpu(i).cpustat.nice); system= cputime64_add(system, kstat_cpu(i).cpustat.system); idle = cputime64_add(idle, kstat_cpu(i).cpustat.idle); iowait = cputime64_add(iowait, kstat_cpu(i).cpustat.iowait); irq = cputime64_add(irq, kstat_cpu(i).cpustat.irq); softirq = cputime64_add(softirq, kstat_cpu(i).cpustat.softirq); steal = cputime64_add(steal, kstat_cpu(i).cpustat.steal); for(j = 0 ; j < NR_IRQS ; j++) sum += kstat_cpu(i).irqs[j]; } …. seq_printf(p, "/nctxt %llu/n" "btime %lu/n" "processes %lu/n" "procs_running %lu/n" "procs_blocked %lu/n", nr_context_switches(), (unsigned long)jif, total_forks, nr_running(), nr_iowait()); …
這部分代碼會輸出你在/proc/stat中看到的數據,通過代碼我們得知iowait來自iowait = cputime64_add(iowait, kstat_cpu(i).cpustat.iowait);
那麼 cpustat.iowait是誰來修改的呢?我們找到了這個函數account_system_time
voidaccount_system_time(structtask_struct *p, inthardirq_offset, cputime_t cputime) { structcpu_usage_stat *cpustat = &kstat_this_cpu.cpustat; structrq *rq = this_rq();//在smp環境下獲取當前的run queue cputime64_t tmp; p->stime = cputime_add(p->stime, cputime); /* Add system time to cpustat. */ tmp = cputime_to_cputime64(cputime); if(hardirq_count() - hardirq_offset)//在做硬中斷 cpustat->irq = cputime64_add(cpustat->irq, tmp); elseif(softirq_count())//在做軟中斷 cpustat->softirq = cputime64_add(cpustat->softirq, tmp); elseif(p != rq->idle)//程序在正常運行,非idle cpustat->system= cputime64_add(cpustat->system, tmp); elseif(atomic_read(&rq->nr_iowait) > 0)//既不做中斷,而且在idle,那麼就是iowait cpustat->iowait = cputime64_add(cpustat->iowait, tmp); else cpustat->idle = cputime64_add(cpustat->idle, tmp); /* Account for system time used */ acct_update_integrals(p); }
我們可以看出,當某個cpu產生iowait時,那麼這個cpu上肯定有進程在進行io,並且在等待io完成(rq->nr_iowait>0),並且這個cpu上沒有進程可運行(p == rq->idle),cpu在idle。
誰在產生iowait?
那麼是誰修改了rq->nr_iowait呢?
重點終於來了,呵呵。
void__sched io_schedule(void) { structrq *rq = &__raw_get_cpu_var(runqueues); delayacct_blkio_start(); atomic_inc(&rq->nr_iowait); schedule(); atomic_dec(&rq->nr_iowait); delayacct_blkio_end(); } long__sched io_schedule_timeout(longtimeout) { structrq *rq = &__raw_get_cpu_var(runqueues); longret; delayacct_blkio_start(); atomic_inc(&rq->nr_iowait); ret = schedule_timeout(timeout); atomic_dec(&rq->nr_iowait); delayacct_blkio_end(); returnret; }
所以產生iowait的根源被我們找到了,就是函數io_schedule, io_schedule_timeout,顧名思義,這兩個函數是用來做進程切換的,而且切換的原因是有io。只不過io_schedule_timeout還給出了一個sleep的時間,也就是timeout。
systemtap來跟一下到底是誰在什麼時候調用了這兩個函數?在這裏我們以引擎爲例子,trace進程searcher_server
Stap腳本Block.stp:(只截取了部分程序)
probe io_schedule = kernel.function("io_schedule"){ #if(tid() == target_pid){ if(isinstr(execname(),"searcher_server")){ stat[execname(), tid(), probefunc()]++; count++; printf("trace time:%s/n", ctime(gettimeofday_s())); print_stack(backtrace()); } } probe io_schedule_timeout = kernel.function("io_schedule_timeout"){ #if(tid() == target_pid){ if(isinstr(execname(),"searcher_server")){ stat[execname(), tid(), probefunc()]++; count++; printf("trace time:%s/n", ctime(gettimeofday_s())); print_stack(backtrace()); } } probe io_schedule, io_schedule_timeout{ } probe begin{ printf("begin %s/n", ctime(gettimeofday_s())); if($# == 2){ if(@1 == "pid") target_pid = strtol(@2, 10) if(@1 == "name") target_name = @2 printf("pid:%d name:%s/n", target_pid, target_name); }else{ printf("arguments error/n"); exit(); } } probe end{ printf("end %s/n", ctime(gettimeofday_s())); } probe timer.ms(1000){ printf("%s running.../n", ctime(gettimeofday_s())); foreach([proc, tid, func] in stat- limit 100){ printf("%s:%d=>%s %d/n", proc, tid, func, stat[proc,tid,func]); } deletestat; printf("%s average schedule times:%d/n", ctime(gettimeofday_s()), count); count = 0; }
程序的大意就是在1S內,統計哪個進程分別調用了多少次這兩個函數。並且把調用時的堆棧print出來,這樣能更清楚地看到到底是哪個系統調用跑到了這個地方。
在最正常的狀態下,跑一下機器:
此時新 引擎 searcher QPS有1500+,cpu busy有88%,iowait幾乎爲0,內存在mmap時全部用MAP_LOCKED被鎖在內存中
sudo stap block.stp pid 5739 -DMAXSKIPPED=1000000 Fri Jul 6 05:57:21 2012 average schedule times:0 Fri Jul 6 05:57:22 2012 running... Fri Jul 6 05:57:22 2012 average schedule times:0 Fri Jul 6 05:57:23 2012 running... Fri Jul 6 05:57:23 2012 average schedule times:0 Fri Jul 6 05:57:24 2012 running... Fri Jul 6 05:57:24 2012 average schedule times:0 Fri Jul 6 05:57:25 2012 running... Fri Jul 6 05:57:25 2012 average schedule times:0 …
跑了一會發現並沒有調用到io schedule,這也符合我們的預期。我們再一邊跑dd一邊stap抓取
sudo stap block.stp pid 5739 -DMAXSKIPPED=1000000 > directdd
起兩個dd進程,寫10G的數據,不走page cache,direct寫
dd if=/dev/zero of=a count=20000000 oflag=direct
dd if=/dev/zero of=b count=20000000 oflag=direct
一共寫20G
-
Searcher表現:
Cpu busy & iowait
Latency:
可以看出direct dd產生的iowait極小,最高才1.4左右,但是對searcher卻也造成了不小的影響,通過vmstat的結果來看,當執行dd之後進程上下文切換從2W+飆到了8W+,被block的searcher線程爲個位數。
-
被block的線程堆棧:
寫log
trace time:Fri Jul 6 06:46:23 2012 0xffffffff8006377c : io_schedule+0x1/0x67 [kernel] 0xffffffff800153e7 : sync_buffer+0x3b/0x3f [kernel] 0xffffffff800639e6 : __wait_on_bit+0x40/0x6e [kernel] 0xffffffff80063a80 : out_of_line_wait_on_bit+0x6c/0x78 [kernel] 0xffffffff8003a9d3 : sync_dirty_buffer+0x96/0xcb [kernel] 0xffffffff8000fd42 : generic_file_buffered_write+0x1cc/0x675 [kernel] 0x00000ffffffff800 trace time:Fri Jul 6 06:46:23 2012 0xffffffff8006377c : io_schedule+0x1/0x67 [kernel] 0xffffffff80028a90 : get_request_wait+0xd8/0x11f [kernel] 0xffffffff8000bfff : __make_request+0x33d/0x401 [kernel] 0xffffffff8001c049 : generic_make_request+0x211/0x228 [kernel] 0xffffffff80033472 : submit_bio+0xe4/0xeb [kernel] 0xffffffff8001a793 : submit_bh+0xf1/0x111 [kernel] 0x00000ffffffff800
-
被block的頻率
grep -E 'run|ks_searcher'directdd | more searcher_server:21001=>io_schedule 1 searcher_server:20813=>io_schedule 1 Fri Jul 6 06:46:20 2012 running... Fri Jul 6 06:46:21 2012 running... Fri Jul 6 06:46:22 2012 running... Fri Jul 6 06:46:23 2012 running... Fri Jul 6 06:46:24 2012 running... searcher_server:20813=>io_schedule 2 searcher_server:21014=>io_schedule 1 Fri Jul 6 06:46:25 2012 running... Fri Jul 6 06:46:26 2012 running... Fri Jul 6 06:46:27 2012 running... Fri Jul 6 06:46:28 2012 running... Fri Jul 6 06:46:29 2012 running...
經stap追查發現,切換次數的增加都是由於direct dd導致的:由於是direct寫,所以每寫一次都要做io schedule
0xffffffff80062391 : schedule+0x1/0xcd4 [kernel] 0xffffffff800637ba : io_schedule+0x3f/0x67 [kernel] 0xffffffff800f281b : __blockdev_direct_IO+0x8bc/0xa35 [kernel] 0xffffffff800c4c91 : generic_file_direct_IO+0xff/0x119 [kernel] 0xffffffff8001edd1 : generic_file_direct_write+0x60/0xf2 [kernel] 0xffffffff8001646e : __generic_file_aio_write_nolock+0x2b8/0x3b6 [kernel]
-
小結:
Searcher latency上升和searcher相對溫和的io schedule、進程切換都有關係,但是這時的主因應該是進程切換,進程切換還會造成頻繁的進程遷移,TLB flush ,Cache pollution。
再做一次新的實驗,把dd的direct flag去掉,讓page cache生效
Searcher的運行環境和運行壓力和上同
-
Searcher表現:
Cpu busy & iowait:
Latency:
可以看出帶page cache的dd對searcher影響更大,我們先看一下vmstat抓取到的數據
平均被block的線程數據很多,甚至在某個時刻可以運行的線程數量爲0
-
searcher被block時的堆棧:
block layer寫請求
trace time:Fri Jul 6 07:13:45 2012 0xffffffff8006377c : io_schedule+0x1/0x67 [kernel] 0xffffffff80028a90 : get_request_wait+0xd8/0x11f [kernel] 0xffffffff8000bfff : __make_request+0x33d/0x401 [kernel] 0xffffffff8001c049 : generic_make_request+0x211/0x228 [kernel] 0xffffffff80033472 : submit_bio+0xe4/0xeb [kernel] 0xffffffff8001a793 : submit_bh+0xf1/0x111 [kernel] 0x00000ffffffff800
Sync buffer
trace time:Fri Jul 6 07:13:46 2012 0xffffffff8006377c : io_schedule+0x1/0x67 [kernel] 0xffffffff800153e7 : sync_buffer+0x3b/0x3f [kernel] 0xffffffff800639e6 : __wait_on_bit+0x40/0x6e [kernel] 0xffffffff80063a80 : out_of_line_wait_on_bit+0x6c/0x78 [kernel] 0xffffffff8003a9d3 : sync_dirty_buffer+0x96/0xcb [kernel] 0xffffffff8001cdc3 : mpage_writepages+0x1bf/0x37e [kernel] 0x00000ffffffff800
此時的dirty ratio已大於40%,需要做blk_congestion_wait,這個可以算是最嚴厲的懲罰了。。
trace time:Fri Jul 6 07:13:48 2012 0xffffffff800631bb : io_schedule_timeout+0x1/0x79 [kernel] 0xffffffff8003b426 : blk_congestion_wait+0x67/0x81 [kernel] 0xffffffff800c7e68 : balance_dirty_pages_ratelimited_nr+0x17d/0x1fa [kernel] 0xffffffff8000fd81 : generic_file_buffered_write+0x20b/0x675 [kernel] 0xffffffff8001651f : __generic_file_aio_write_nolock+0x369/0x3b6 [kernel] 0xffffffff8002157b : generic_file_aio_write+0x65/0xc1 [kernel] 0x00000ffffffff800
Searcher用到的某些頁被刷出去,需要sync page read
trace time:Fri Jul 6 07:13:49 2012 0xffffffff8006377c : io_schedule+0x1/0x67 [kernel] 0xffffffff80028936 : sync_page+0x3e/0x43 [kernel] 0xffffffff800638fe : __wait_on_bit_lock+0x36/0x66 [kernel] 0xffffffff8003fbad : __lock_page+0x5e/0x64 [kernel] 0xffffffff800139f8 : find_lock_page+0x69/0xa2 [kernel] 0xffffffff800c45a5 : grab_cache_page_write_begin+0x2c/0x89 [kernel] 0x00000ffffffff800
-
被block的頻率:
searcher_server:21010=>io_schedule 3 searcher_server:21003=>io_schedule 1 Fri Jul 6 07:14:39 2012 running... searcher_server:21010=>io_schedule 7 Fri Jul 6 07:14:40 2012 running... searcher_server:21008=>io_schedule 1 Fri Jul 6 07:14:41 2012 running... Fri Jul 6 07:14:42 2012 running... searcher_server:21004=>io_schedule 1 Fri Jul 6 07:14:43 2012 running... searcher_server:21014=>io_schedule 11 searcher_server:21015=>io_schedule 1 searcher_server:21008=>io_schedule 1 Fri Jul 6 07:14:44 2012 running... Fri Jul 6 07:14:45 2012 running... Fri Jul 6 07:14:46 2012 running... searcher_server:21003=>io_schedule 2 Fri Jul 6 07:14:47 2012 running... Fri Jul 6 07:14:48 2012 running... Fri Jul 6 07:14:49 2012 running...
-
小結:
當帶page cahce進行dd時,很容易就能達到10%的background dirty ratio和40%的dirty ratio,達到40%的時候buffered write就變成了sync write。經stap trace發現每次blk_congestion_wait都要耗時100ms左右,也就是說一個線程要被block 100ms,很致命。
爲了減少io的影響,我們把log給禁掉
再做一次實驗
-
Searcher表現:
Cpu busy & iowait
Latency:
把寫log關掉之後竟然還有iowait,是誰造成的呢?
-
Searcher被block時的堆棧:
trace time:Fri Jul 6 02:08:28 2012 0xffffffff8006377c : io_schedule+0x1/0x67 [kernel] 0xffffffff80028936 : sync_page+0x3e/0x43 [kernel] 0xffffffff800638fe : __wait_on_bit_lock+0x36/0x66 [kernel] 0xffffffff8003fbad : __lock_page+0x5e/0x64 [kernel] 0xffffffff80013881 : filemap_nopage+0x268/0x360 [kernel] 0xffffffff8000898c : __handle_mm_fault+0x1fa/0xf99 [kernel] 0x00000ffffffff800
我們的內存都被mlock了,竟然還有sync page,爲啥呢?
用blktrace和debugfs追了一下,發現竟然是一個算法數據的問題
/path/of/data
原來是這個文件的數據被dd給刷出去了,導致還要重新read到內存
然後寫了個程序把這個數據也lock到內存中
./lock /path/of/data
Lock數據再重新跑dd with page cache
-
Searcher 表現:
Cpu busy & iowait:
Latency:
可以看到,iowait水平又降低了不少,那麼此時此刻,誰還在製造iowait呢?
-
Searcher被block時的堆棧:
trace time:Fri Jul 6 03:45:04 2012 0xffffffff800631bb : io_schedule_timeout+0x1/0x79 [kernel] 0xffffffff8003b426 : blk_congestion_wait+0x67/0x81 [kernel] 0xffffffff800ca438 : try_to_free_pages+0x252/0x2d7 [kernel] 0xffffffff8000f40d : __alloc_pages+0x1cb/0x2ce [kernel] 0xffffffff80008e62 : __handle_mm_fault+0x6d0/0xf99 [kernel] 0xffffffff80066b25 : do_page_fault+0x4cb/0x830 [kernel] trace time:Fri Jul 6 03:45:04 2012 0xffffffff800631bb : io_schedule_timeout+0x1/0x79 [kernel] 0xffffffff8003b426 : blk_congestion_wait+0x67/0x81 [kernel] 0xffffffff800ca438 : try_to_free_pages+0x252/0x2d7 [kernel] 0xffffffff8000f40d : __alloc_pages+0x1cb/0x2ce [kernel] 0xffffffff8002600b : tcp_sendmsg+0x567/0xb0e [kernel] 0xffffffff80037c60 : do_sock_write+0xc6/0x102 [kernel]
原來是內存很少了,導致申請內存時要走到try_to_free_pages(平時極少走到),走到這一步說明系統內存已經少的可憐。但是沒辦法,誰讓searcher還要去malloc呢,這些malloc來自兩部分:1,mempool申請的內存,其實這個是完全可以抹掉的 2,算法so中STL部分用到的內存。
-
小結:
關掉log,將數據都lock在內存中,降低了iowait的水平,但是要讓searcher不受影響,還要做更多的工作,比如不申請內存。
如何消除searcher(或應用系統)的iowait?
1,沒有io
不寫log,或者把寫log的事情交給一個專門線程來做,searcher不做buffered write;不做disk read,尤其是sync page這類操作。
2,全內存且不申請內存
用到的數據read once,全內存且lock住;把mempool做到完美,起碼做到99%的case不做內存申請。
3,儘量減少其他應用的io影響
其實就是能將dd的負面影響降到最少,如用cgroup;在scp多個大文件的時候,在傳輸過程中及時清理每個大文件的page cache,將系統的dirty ratio維持在10%以下,尤其是不能達到40%。