From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sergei Trofimovich Subject: Re: Rename BTRfs to MuchSlowerFS ? Date: Mon, 5 Sep 2011 22:25:23 +0300 Message-ID: <20110905222523.4531abe0@sf.home> References: <4E64D3D5.7020407@petaramesh.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=PGP-SHA1; boundary="Sig_/.nLyE1A5k3chceS87LJ4Cmq"; protocol="application/pgp-signature" Cc: linux-btrfs To: =?UTF-8?B?U3fDom1p?= Petaramesh Return-path: In-Reply-To: <4E64D3D5.7020407@petaramesh.org> List-ID: --Sig_/.nLyE1A5k3chceS87LJ4Cmq Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable On Mon, 05 Sep 2011 15:51:17 +0200 Sw=C3=A2mi Petaramesh wrote: > Hi list, >=20 > I don't trust theoretical benchmarks that much and prefer "real-life=20 > benchs" on the occasion, so here's mine: >=20 > Given 4 laptops, the most powerful of which was running BTRFS and the=20 > others ext3 or ext4, all machines running Ubuntu 11.04 Natty 32-bit with= =20 > a stock Ubuntu 2.6.38-11 kernel, all machines were given the following=20 > FS-intensive task : >=20 > - Upgrade O.S. from Ubuntu 11.04 Natty to 11.10 Oneiric (beta), using a=20 > local packages apt-cacher. >=20 > Machines : >=20 > 1/ Acer Aspire 3104 WLMi, AMD Sempron Mobile 3500+ @1.8 GHz, 1.5 GB RAM,= =20 > 80 GB SATA HD, ext4 over standard partitions >=20 > 2/ Asus EeePC 1005PE, Intel Atom N450 @1.66 GHz, 2 GB RAM, 250 GB SATA=20 > HD, ext3 over AES-128 encrypted LVM >=20 > 3/ Compaq Mini CQ10-740SF, Intel Atom N455 @1.66 GHz, 1 GB RAM, 250 GB=20 > SATA HD, ext4 over standard partitions >=20 > 4/ Dell XPS M1330, Intel Core 2 Duo T9300 @2.5 GHz, 2 GB RAM, 160 B SATA= =20 > HD, BTRFS over AES-128 encrypted LVM >=20 > Results : >=20 > All 3 ext3 / ext4 machines took between 60 and 90 minutes to complete=20 > their upgrade. >=20 > BTRFS machine took 20 HOURS so far, still counting (ETA 15 minutes left). >=20 > Wow. Impressive. I've seen similar problem on Ubuntu-11 + Aspire One (8GB of slow "SSD"). More specifically half of ubuntu install went very fast and when disk was ~50% free things _suddenly_ gone slow. According to strace fsync() started to take ~10 seconds(!) and upper. It's not common filesystem behaviour, but a plain (and hard to catch?) bug. I don't have the box anymore. The root cause it to be found by some brave soul yet. I wonder if latencytop would get the profile of the issue. Maybe, btrfs degrades only after certain (insanely large?) amount of fsync() calls? --=20 Sergei --Sig_/.nLyE1A5k3chceS87LJ4Cmq Content-Type: application/pgp-signature; name=signature.asc Content-Disposition: attachment; filename=signature.asc -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.17 (GNU/Linux) iEYEARECAAYFAk5lIigACgkQcaHudmEf86oXuwCeLRXbXyNamINqcYTDDlLqjqSY EpgAn1tk6mGiGuzeR5+9hnZm0mDGY2bg =Z9vu -----END PGP SIGNATURE----- --Sig_/.nLyE1A5k3chceS87LJ4Cmq--