Does sync(2) block writes until completed?


 
Thread Tools Search this Thread
Top Forums UNIX for Advanced & Expert Users Does sync(2) block writes until completed?
# 1  
Old 09-24-2008
Does sync(2) block writes until completed?

Gentle readers,
I am trying to observe system behavior on our RHEL 5.2 machines. I notice that, it appears to me, based on random iterations of
Code:
dd if=/dev/zero of=/var/tmp/bigfile bs=1024 count=20000

...that if the pdflush is flushing buffers at the time of my write, my write will take about as long as
Code:
dd if=/dev/zero of=/var/tmp/bigfile bs=1024 count=20000 ; sync

This indicates to me that pdflush blocks all write()'s until the flush to disk is complete. The question is: am I correct?

Most of the time, the first dd (without the sync) is much faster than the second. By watching /proc/meminfo, I am able to see approximately when the sync initiated by pdflush takes place. This would be every 30 seconds, as my machine is not filling the disk buffers very quickly. We're talking only a few hundred kilobytes every 30 seconds.

Thanks.
-Mike
# 2  
Old 09-24-2008
When sync is called - maybe by the syncer daemon, it requests the kernel to write all of it's buffers from all buffers for all file descriptors to disk. The kernel does not have to do it exactly at that point, it is a scheduled request.

sync returns before the completion of all the kernel I/O it asked to schedule. It can return before any I/O starts in fact.

On systems with no buffer cache, sync may just be a dummy call that does nothing, because no 'syncing' is required. On a multiprocessor system, processes may continue to call the write() system call when you call sync -- it will not block writes.

I don't know what you are seeing, but sync is not necessarily the cause. try
Code:
 strace -rT sync

and see what you get.
# 3  
Old 09-24-2008
Not sync, really- but pdflush

I should mention that I'm not really concerned so much about sync but rather on pdflush itself. ...The mechanisms they use may be different. Anyway, I notice that if I modify the pdflush behavior with these kernel parameters:
Code:
vm.dirty_writeback_centisecs=100
vm.dirty_expire_centisecs=100

...it appears to affect my dd writes negatively; that is, if I loop the dd and run it over and over again, the time for dd to complete jumps on an erratic basis.

Here's what I'm executing:
Code:
while true; do
    echo -n ">>> New File >>>"; date +%r
    /usr/bin/time -f "%E" /tmp/doit
done

and /tmp/doit contains:
dd if=/dev/zero of=/var/tmp/bigfile$i bs=1024 count=40000 oflag=nonblock 2>/dev/null

Then in another window, I am modifying sysctl parameters while the while loop is running. First, the default settings:
Code:
sysctl -w vm.dirty_writeback_centisecs=500 ; sysctl -w vm.dirty_expire_centisecs=3000
sysctl -w vm.dirty_writeback_centisecs=100 ; sysctl -w vm.dirty_expire_centisecs=100
...observe times from the while loop.  These should be more "choppy"

The fact that the dd's more frequently take longer when the centisecs parameter is shortened lead me to believe that my app is blocking when it bumps into the disk flushing. The machine is otherwise quiescent, and the size of the write (40 Meg) should not be enough to trigger pdflush. It should happen every second, based on the sysctl settings.

Note that dd is not doing direct i/o or synchronous i/o. If I explicitly set those flags, it slows the write enormously. Note too that it doesn't matter if I use the "nonblock" flag or not... i/o behavior is the same.
-Mike
P.S. We don't have a syncer daemon. Also, while running my tests above, it appears to me that the effect of sync is nearly instantaneous...

Last edited by mschwage; 09-24-2008 at 05:42 PM..
# 4  
Old 09-24-2008
Here is a discussion on pdflush and how a sync call can bog the system - read down about 5 paragraphs in the Tuning... section

The Linux Page Cache and pdflush
# 5  
Old 09-24-2008
Thanks, Jim, but I've been all over that page. The one part that gave me hope is
Quote:
processes are forced to write dirty buffers themselves...Note that all processes are blocked for writes when this happens, not just the one that filled the write buffers.
but then I realized that it didn't apply. It says that the processes will write during this anomalous period, but what I'm trying to get at is normal everyday pdflush behavior. I wonder if pdflush blocks other processes during their write(2)'ing until the blocks are flushed to disk.

...Oh, I see... well, I've been down to the Tuning section, and it seems to imply that pdflush blocks, but I'm still not sure...
-Mike
Login or Register to Ask a Question

Previous Thread | Next Thread

8 More Discussions You Might Find Interesting

1. Solaris

OpenBSM not catching all file writes

I have a custom auditing class configured for these events but it doesn't seem to be catching it when I do a "echo hey > test.txt" Any ideas on why that is? ---------- Post updated at 06:04 PM ---------- Previous update was at 05:47 PM ---------- Scratch this topic, it looks like it's because... (0 Replies)
Discussion started by: thmnetwork
0 Replies

2. Shell Programming and Scripting

Help with script that reads and writes java console Minecraft

Hi I am looking for an easy way to lock game mode (0) for everyone included op on a Minecraft server. It can be a script that every time a player changes game to 1 the script changes back to 0. What the player writes is visible in the java console. I am not good at script programming and my... (0 Replies)
Discussion started by: MyMorris
0 Replies

3. Shell Programming and Scripting

Reading from a file a background program writes to

Hi! #!/usr/bin/env bash rm tmpcomm nc -v -u -l 444 | hexdump -b > tmpcomm while : do read l1 < tmpcomm read l2 < tmpcomm read l3 < tmpcomm read l4 < tmpcomm # do something doneI start netcat in the background and listen for an incoming conncetion. All incoming... (1 Reply)
Discussion started by: torax123
1 Replies

4. Solaris

Network writes contantly spiking in throughput

Hey guys First post... and im not exactly a solaris guru but here goes Ive setup a solaris 10 box with a raidz2 set of 6 disks... I have also setup Samba with open shares for some CIFs access... now my issue is that when i transfer large files to it the network performance contantly... (8 Replies)
Discussion started by: silicoon
8 Replies

5. Shell Programming and Scripting

Perl: FH and multiple writes

I found this logging subroutine on the net that I want to use but despite trying many things, I cannot figure out how to get the date in front of the logged text. Ideally what I'm looking for is a line that looks something like this: Wed Aug 20 18:17:29 PDT 2008 - my logging info here. my... (2 Replies)
Discussion started by: gctaylor
2 Replies

6. UNIX for Dummies Questions & Answers

which process writes to file

Some process rewrites a file ( i'm hacked :) Can I somehow monitor which process does that? (2 Replies)
Discussion started by: hachik
2 Replies

7. Solaris

Sync to Green vs. Separate Sync

Hi all....I have a Sun Ultra2 that I want to use with my PC monitor. I have purchased an adapter that does not work and I was told I need to change my video card setting (if I can) to Separate Sync.....my Monitor product number ends in 1343......I am running SunOS 5.7 ......anyone have any ideas? ... (0 Replies)
Discussion started by: psantinello
0 Replies

8. Shell Programming and Scripting

sh script that reads/writes based upon contents of a file

Hi everyone, Ive got a quick question about the feasibility and any suggestions for a shell script. I can use sh or ksh, doesnt matter. Basically, Ive got an output file from a db2 command that looks like so: SCHEMA NAME CARD LEAF ELEAF LVLS ISIZE NDEL KEYS F4 F5 ... (3 Replies)
Discussion started by: rdudejr
3 Replies
Login or Register to Ask a Question