ext4 numbers on my ubuntu 11.10 don't look all that good. The write is blocked significantly
rite in 41 microseconds Write in 25 microseconds Sync in 2803 microseconds (0) Write in 858 microseconds Write in 27 microseconds Write in 21 microseconds Write in 24 microseconds Write in 26 microseconds Write in 26 microseconds Write in 23 microseconds Write in 27 microseconds Write in 20 microseconds Write in 37 microseconds Sync in 47403 microseconds (0) Write in 45376 microseconds Write in 27 microseconds Write in 23 microseconds Write in 45 microseconds Write in 29 microseconds Write in 50 microseconds Write in 29 microseconds Write in 26 microseconds Write in 26 microseconds Write in 28 microseconds Sync in 36169 microseconds (0) Write in 32379 microseconds Write in 27 microseconds Write in 51 microseconds Write in 44 microseconds Write in 26 microseconds Write in 44 microseconds Write in 26 microseconds Write in 28 microseconds Write in 25 microseconds Write in 27 microseconds Sync in 3356 microseconds (0) Write in 1888 microseconds Write in 23 microseconds Write in 45 microseconds Write in 25 microseconds Write in 44 microseconds Write in 50 microseconds Write in 40 microseconds Write in 35 microseconds Write in 104 microseconds Thanks! -neelesh On Thu, May 24, 2012 at 3:01 PM, Neelesh <neele...@gmail.com> wrote: > one of the comments in the article you mentioned has some numbers that > essentially say ext4 behaves much better (i haven't tested this yet, I'll > do it tonight and post the results). > > thanks > -neelesh > > > On Thu, May 24, 2012 at 10:40 AM, Jay Kreps <jay.kr...@gmail.com> wrote: > >> One issue with using the filesystem for persistence is that the >> synchronization in the filesystem is not great. In particular the fsync >> and >> fsyncdata system calls block appends to the file, apparently for the >> entire >> duration of the fsync (which can be quite long). This is documented in >> some >> detail here: >> http://antirez.com/post/fsync-different-thread-useless.html >> >> This is a problem in 0.7 because our definition of a committed message is >> one written prior to calling fsync(). This is the only way to guarantee >> the >> message is on disk. We do not hand out any messages to consumers until an >> fsync call occurs. The problem is that regardless of whether the fsync is >> in a background thread or not it will block any produce requests to the >> file. This is buffered a bit in the client since our produce request is >> effectively async in 0.7, but it can lead to weird latency spikes >> nontheless as this buffering gets filled. >> >> In 0.8 with replication the definition of a committed message changes to >> one that is replicated to multiple machines, not necessarily committed to >> disk. This is a different kind of guarantee with different strengths and >> weaknesses (pro: data can survive destruction of the file system on one >> machine, con: you will lose a few messages if you haven't sync'd and the >> power goes out). We will likely retain the flush interval and time >> settings >> for those who want fine grained control over flushing, but it is less >> relevant. >> >> Unfortunately *any* call to fsync will block appends even in a background >> thread so how can we give control over physical disk persistence without >> introducing high latency for the producer? The answer is that the linux >> pdflush daemon actually does a very similar thing to our flush parameters. >> pdflush is a daemon running on every linux machine that controls the >> writing of buffered/cached data back to disk. It allows you to control the >> percentage of memory filled with dirty pages by giving it either a >> percentage of memory, a time out for any dirty page to be written, or a >> fixed number of dirty bytes. >> >> The question is, does pdflush block appends? The answer seems to be mostly >> no. It locks the page being flushed but not the whole file. The time to >> flush one page is actually usually pretty quick (plus I think it may not >> be >> flushing just written pages anyway). I wrote some test code for this and >> here are the results: >> >> I modified the code from the link above. Here are the results from my >> desktop (Centos Linux 2.6.32). >> >> We run the test writing 1024 bytes every 100 us and flushing every 500 us: >> >> $ ./pdflush-test 1024 100 500 >> 21 >> 4 >> 3 >> 3 >> 9 >> 6 >> Sync in 20277 us (0), sleeping for 500 us >> 19819 >> 7 >> 7 >> 8 >> 38 >> Sync in 19470 us (0), sleeping for 500 us >> 19048 >> 7 >> 4 >> 3 >> 8 >> 4 >> Sync in 19405 us (0), sleeping for 500 us >> 19017 >> 6 >> 6 >> 10 >> 6 >> Sync in 19410 us (0), sleeping for 500 us >> 19025 >> 7 >> 7 >> 11 >> 6 >> >> $ cat /proc/sys/vm/dirty_writeback_centisecs >> 100 >> $ cat /proc/sys/vm/dirty_expire_centisecs >> 500 >> >> Now run the test with the background flush disabled (rarely running): >> $ ./pdflush-test 1024 100 5000000000000 > times.txt >> >> I ran this for 298,028 writes. The 99.9th percentile for this test is 17 >> us >> and the max time was 2043 us (2ms). >> >> Here is the test code: >> >> #include <stdio.h> >> #include <unistd.h> >> #include <string.h> >> #include <sys/types.h> >> #include <pthread.h> >> #include <sys/stat.h> >> #include <fcntl.h> >> #include <sys/time.h> >> #include <stdlib.h> >> >> static long long microseconds(void) { >> struct timeval tv; >> long long mst; >> >> gettimeofday(&tv, NULL); >> mst = ((long long)tv.tv_sec)*1000000; >> mst += tv.tv_usec; >> return mst; >> } >> >> void *IOThreadEntryPoint(void *arg) { >> int fd, retval; >> long long start; >> long sleep = (long) arg; >> >> while(1) { >> usleep(sleep); >> start = microseconds(); >> fd = open("/tmp/foo.txt",O_RDONLY); >> retval = fsync(fd); >> close(fd); >> printf("Sync in %lld us (%d), sleeping for %ld us\n", >> microseconds()-start, retval, sleep); >> } >> return NULL; >> } >> >> int main(int argc, char* argv[]) { >> if(argc != 4) { >> printf("USAGE: %s size write_sleep fsync_sleep\n", argv[0]); >> exit(1); >> } >> >> pthread_t thread; >> int fd = open("/tmp/foo.txt",O_WRONLY|O_CREAT,0644); >> long long start; >> long long ellapsed; >> int size = atoi(argv[1]); >> long write_sleep = atol(argv[2]); >> long fsync_sleep = atol(argv[3]); >> char buff[size]; >> >> pthread_create(&thread,NULL,IOThreadEntryPoint, (void*) fsync_sleep); >> >> while(1) { >> start = microseconds(); >> if (write(fd,buff,size) == -1) { >> perror("write"); >> exit(1); >> } >> ellapsed = microseconds()-start; >> printf("%lld\n", ellapsed); >> usleep(write_sleep); >> } >> close(fd); >> exit(0); >> } >> >> Cheers, >> >> -Jay >> > >
