Received: by 10.224.186.20 with SMTP id cq20mr9320315qab.8.1352871740451;
        Tue, 13 Nov 2012 21:42:20 -0800 (PST)
X-BeenThere: mongodb-user@googlegroups.com
Received: by 10.224.97.129 with SMTP id l1ls270008qan.6.gmail; Tue, 13 Nov
 2012 21:41:52 -0800 (PST)
Received: by 10.224.105.205 with SMTP id u13mr9308848qao.6.1352871712253;
        Tue, 13 Nov 2012 21:41:52 -0800 (PST)
Received: by 10.224.16.202 with SMTP id p10msqaa;
        Tue, 13 Nov 2012 21:38:02 -0800 (PST)
Received: by 10.52.72.133 with SMTP id d5mr56296vdv.16.1352871482251;
        Tue, 13 Nov 2012 21:38:02 -0800 (PST)
Date: Tue, 13 Nov 2012 21:38:01 -0800 (PST)
From: Gevorg Hari <gevorgh...@gmail.com>
To: mongodb-user@googlegroups.com
Message-Id: <dc7bb4a3-8362-4ed5-9c06-978cb84332e5@googlegroups.com>
In-Reply-To: <a32cfea7-b9a6-42b0-bf5b-dc0ddea6b963@googlegroups.com>
References: <a32cfea7-b9a6-42b0-bf5b-dc0ddea6b963@googlegroups.com>
Subject: Re: DataFileSync - every 30 mins I/O write 100%
MIME-Version: 1.0
Content-Type: multipart/mixed; 
	boundary="----=_Part_401_18236834.1352871481627"

------=_Part_401_18236834.1352871481627
Content-Type: multipart/alternative; 
	boundary="----=_Part_402_20294120.1352871481627"

------=_Part_402_20294120.1352871481627
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit

I'm having a similar issue (no solution yet). The same message appears (not 
always) at intervals of about 1 minute. The time that it takes to flush 
mmaps is usually around 10000ms for  150 files (~300GB) but I've seen the 
flushing taking more than 100000ms at times. During this period MongoDB is 
locked. No queries can be executed (not even simple ones from a terminal on 
the same host) and I've been disconnected a couple of times from a python 
script. I'm running MongoDB 2.2.1 for Windows 64-bit 2008R2+ on a WIndows 
Server 2008 R2. Help, I'm starting not to sleep well at night...

On Sunday, November 4, 2012 1:45:06 PM UTC-5, Anton Swart wrote:
>
> Every 30 minutes mongod 2.2.1 on Windows is causing 100% write IO to disk.
>
> The log file containt following:
>
> Sun Nov 04 16:31:49 [DataFileSync] flushing mmaps took 73578ms  for 12 
> files
> Sun Nov 04 16:33:19 [DataFileSync] flushing mmaps took 29704ms  for 12 
> files
>
> Sun Nov 04 17:01:02 [DataFileSync] flushing mmaps took 12782ms  for 12 
> files
> Sun Nov 04 17:31:02 [DataFileSync] flushing mmaps took 13469ms  for 12 
> files
>
> Sun Nov 04 20:32:12 [DataFileSync] flushing mmaps took 48563ms  for 11 
> files
> Sun Nov 04 20:33:01 [DataFileSync] flushing mmaps took 37844ms  for 11 
> files
>
> etc every 30 mins. This is killing me!
>
>
> I have one DB with 4 files totalling 1.9GB
> Weirdly mongo is only using < 200mb of RAM
>
> Also sitting consistently with 4000 page faults.
>
> I installed MMS so can provide more info if needed.
>
> Please help!
>
>
------=_Part_402_20294120.1352871481627
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable

I'm having a similar issue (no solution yet). The same message appears (not=
 always) at intervals of about 1 minute. The time that it takes to flush mm=
aps is usually around 10000ms for &nbsp;150 files (~300GB) but I've seen th=
e flushing taking more than 100000ms at times. During this period MongoDB i=
s locked. No queries can be executed (not even simple ones from a terminal =
on the same host) and I've been disconnected a couple of times from a pytho=
n script. I'm running MongoDB 2.2.1 for Windows 64-bit 2008R2+ on a WIndows=
 Server 2008 R2. Help, I'm starting not to sleep well at night...<br><br>On=
 Sunday, November 4, 2012 1:45:06 PM UTC-5, Anton Swart wrote:<blockquote c=
lass=3D"gmail_quote" style=3D"margin: 0;margin-left: 0.8ex;border-left: 1px=
 #ccc solid;padding-left: 1ex;"><div>Every 30 minutes mongod 2.2.1 on Windo=
ws is causing 100% write IO to disk.</div><div><br></div><div>The log file =
containt following:</div><div><br></div><div>Sun Nov 04 16:31:49 [DataFileS=
ync] flushing mmaps took 73578ms &nbsp;for 12 files<br></div><div>Sun Nov 0=
4 16:33:19 [DataFileSync] flushing mmaps took 29704ms &nbsp;for 12 files<br=
></div><div><br></div><div>Sun Nov 04 17:01:02 [DataFileSync] flushing mmap=
s took 12782ms &nbsp;for 12 files<br></div><div>Sun Nov 04 17:31:02 [DataFi=
leSync] flushing mmaps took 13469ms &nbsp;for 12 files<br></div><div><br></=
div><div>Sun Nov 04 20:32:12 [DataFileSync] flushing mmaps took 48563ms &nb=
sp;for 11 files<br></div><div>Sun Nov 04 20:33:01 [DataFileSync] flushing m=
maps took 37844ms &nbsp;for 11 files<br></div><div><br></div><div>etc every=
 30 mins. This is killing me!</div><div><br></div><div><br></div><div>I hav=
e one DB with 4 files totalling 1.9GB</div><div>Weirdly mongo is only using=
 &lt; 200mb of RAM</div><div><br></div><div>Also sitting consistently with =
4000 page faults.</div><div><br></div><div>I installed MMS so can provide m=
ore info if needed.</div><div><br></div><div>Please help!</div><div><br></d=
iv></blockquote>
------=_Part_402_20294120.1352871481627--

------=_Part_401_18236834.1352871481627--