lkml.org 
[lkml]   [2008]   [Dec]   [28]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: [0/7] Distributed storage release.
Date
Pavel Machek wrote:
> On Fri 2008-12-26 14:56:09, Evgeniy Polyakov wrote:
>> Hi.
>>
>> This is a maintenance distributed storage release, which includes a
>> rebase against the 2.6.28 kernel tree only.
>>
>> DST is a network block device storage, which can be used to organize
>> exported storages on the remote nodes into the local block device.
>>
>> Its main goal of the project is to allow creation of the block devices
>> on top of different network media and connect physically distributed devices
>> into single storage using existing network infrastructure and not
>> introducing new limitations into the protocol and network usage model.
>
> So it is basically nbd on steroids?
>
> ...reminds me, nbd-server should really fsync data before returning success...
> Pavel
>
If you really want reliable operation without killing performance on the server
you could use aio and just wait until the data is written to the device. Lots of
discussion of this in various places, "on the device" may not mean on the
platter, then you talk disabling write cache, barriers, etc. At least aio should
(a) be more reliable than just issuing the i/o and (b) not impact the
performance of the server as fsync would.

Comments welcome, pointers to a benchmark even more so, now I'm curious.

--
Bill Davidsen <davidsen@tmr.com>
"We have more to fear from the bungling of the incompetent than from
the machinations of the wicked." - from Slashdot



\
 
 \ /
  Last update: 2008-12-28 19:21    [from the cache]
©2003-2011 Jasper Spaans. Advertise on this site