xfs
[Top] [All Lists]

Re: Pathological allocation pattern with direct IO

To: Linux fs XFS <xfs@xxxxxxxxxxx>
Subject: Re: Pathological allocation pattern with direct IO
From: pg_xf2@xxxxxxxxxxxxxxxxxx (Peter Grandi)
Date: Wed, 6 Mar 2013 22:31:42 +0000
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <20130306202210.GA1318@xxxxxxxxxxxxx>
References: <20130306202210.GA1318@xxxxxxxxxxxxx>
Resent-date: Thu, 7 Mar 2013 21:04:30 +0000
Resent-from: pg_mh@xxxxxxxxxxxxxxxx
Resent-message-id: <20793.222.376461.229641@xxxxxxxxxxxxxxxxxx>
Resent-to: xfs@xxxxxxxxxxx
> [ ... ] application that write large (tens of GB) files using
> direct IO done in 16 MB chunks. They keep the fs around 80%
> full deleting oldest files when they need to store new ones.
> Usually the file can be stored in under 10 extents but from
> time to time a pathological case is triggered and the file has
> few thousands extents (which naturally has impact on
> performance). [ ... ]

They just want *guaranteed* (not just nearly-always) contiguity,
without preallocating, with incremental writes with direct IO,
while keeping the filesystem full at 80%.

And a pony! :-)

<Prev in Thread] Current Thread [Next in Thread>