Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.os.linux.misc > #11848
| From | Pascal Hambourg <boite-a-spam@plouf.fr.eu.org> |
|---|---|
| Newsgroups | comp.os.linux.misc |
| Subject | Re: External hard drive usage. NTFS or other file system? |
| Date | 2014-08-30 14:42 +0200 |
| Organization | Plouf ! |
| Message-ID | <ltsgrt$1nge$1@saria.nerim.net> (permalink) |
| References | (1 earlier) <lrdv9a$ma1$1@dont-email.me> <pan.2014.08.29.02.46.26@gmail.com> <ltor78$vgo$1@dont-email.me> <6pc6db-8gi.ln1@user-1943a.linux.individual.net> <pan.2014.08.29.15.19.35@gmail.com> |
Mladen Gogala a écrit : > > The reason for defragmenting a drive are virtual machines, of which I > have quite a few. Since I am playing with thing like Oracle RAC, I > frequently have to run 3 or 4 of them simultaneously. They perform > visibly better if VMDK files are defragmented. The reason is simple to > explain: a database full scan, which does a sequential scan on a virtual > machine, will also do a sequential scan on the real drive if the > underlying VMDK file is contiguous. If not, it will be translated into > gazillion of small fragment scans, which can take time. There is a huge gap between "contiguous" and "gazillion of small fragments". If fragments are big enough (wrt. the disk sequential speed * access time product), the adverse effects of fragmentation should be minimized. In ext4, isn't the use of extents supposed to reduce fragmentation and improve large file performance ? E.g. : Typical hard disk with 10 ms average access type and 100 MB/s sequential speed. 100 MB/s * 10 ms = 1 MB. Extents in ext4 can map up to 128 MiB of contiguous space, which is two orders of magnitude bigger. Besides, shouldn't VM managers be aware of this issue and grow virtual disk files by big chunks in order to limit fragmentation too ?
Back to comp.os.linux.misc | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
External hard drive usage. NTFS or other file system? Evgenii Sputnik <esputnik7788@gmail.com> - 2014-07-31 13:35 -0400
Re: External hard drive usage. NTFS or other file system? Aragorn <thorongil@telenet.be.invalid> - 2014-07-31 19:44 +0200
Re: External hard drive usage. NTFS or other file system? Evgenii Sputnik <esputnik7788@gmail.com> - 2014-08-29 07:45 +0700
Re: External hard drive usage. NTFS or other file system? Aragorn <thorongil@telenet.be.invalid> - 2014-08-29 03:37 +0200
Re: External hard drive usage. NTFS or other file system? Robert Heller <heller@deepsoft.com> - 2014-08-28 21:09 -0500
Re: External hard drive usage. NTFS or other file system? David Brown <david.brown@hesbynett.no> - 2014-08-29 21:16 +0200
Re: External hard drive usage. NTFS or other file system? Mladen Gogala <gogala.mladen@gmail.com> - 2014-08-29 02:46 +0000
Re: External hard drive usage. NTFS or other file system? Aragorn <thorongil@telenet.be.invalid> - 2014-08-29 05:14 +0200
Re: External hard drive usage. NTFS or other file system? William Poaster <wp@dev.null> - 2014-08-29 14:45 +0100
Re: External hard drive usage. NTFS or other file system? Mladen Gogala <gogala.mladen@gmail.com> - 2014-08-29 15:19 +0000
Re: External hard drive usage. NTFS or other file system? Pascal Hambourg <boite-a-spam@plouf.fr.eu.org> - 2014-08-30 14:42 +0200
Re: External hard drive usage. NTFS or other file system? Robert Heller <heller@deepsoft.com> - 2014-08-30 09:09 -0500
Re: External hard drive usage. NTFS or other file system? Pascal Hambourg <boite-a-spam@plouf.fr.eu.org> - 2014-08-30 17:41 +0200
Re: External hard drive usage. NTFS or other file system? Keith Keller <kkeller-usenet@wombat.san-francisco.ca.us> - 2014-08-30 15:21 -0700
Re: External hard drive usage. NTFS or other file system? Mladen Gogala <gogala.mladen@gmail.com> - 2014-08-31 17:13 +0000
Re: External hard drive usage. NTFS or other file system? Robert Heller <heller@deepsoft.com> - 2014-08-31 13:24 -0500
Is VirtualBox "lame"? (Was: something else...) gazelle@shell.xmission.com (Kenny McCormack) - 2014-08-31 20:23 +0000
Re: Is VirtualBox "lame"? (Was: something else...) Robert Heller <heller@deepsoft.com> - 2014-08-31 15:51 -0500
Re: Is VirtualBox "lame"? Rich <rich@example.invalid> - 2014-08-31 23:58 +0000
Re: Is VirtualBox "lame"? Robert Heller <heller@deepsoft.com> - 2014-08-31 21:54 -0500
Re: Is VirtualBox "lame"? The Natural Philosopher <tnp@invalid.invalid> - 2014-09-01 07:54 +0100
Re: Is VirtualBox "lame"? Marc Haber <mh+usenetspam1118@zugschl.us> - 2014-09-01 08:53 +0200
Re: Is VirtualBox "lame"? (Was: something else...) The Natural Philosopher <tnp@invalid.invalid> - 2014-09-01 07:47 +0100
Re: External hard drive usage. NTFS or other file system? Mladen Gogala <gogala.mladen@gmail.com> - 2014-08-31 21:00 +0000
Re: External hard drive usage. NTFS or other file system? Rich <rich@example.invalid> - 2014-08-31 18:50 +0000
Re: External hard drive usage. NTFS or other file system? Mladen Gogala <gogala.mladen@gmail.com> - 2014-08-31 21:01 +0000
Re: External hard drive usage. NTFS or other file system? Rich <rich@example.invalid> - 2014-09-01 00:06 +0000
Re: External hard drive usage. NTFS or other file system? The Natural Philosopher <tnp@invalid.invalid> - 2014-09-01 07:50 +0100
Re: External hard drive usage. NTFS or other file system? Mladen Gogala <gogala.mladen@gmail.com> - 2014-08-31 03:18 +0000
Re: External hard drive usage. NTFS or other file system? Robert Heller <heller@deepsoft.com> - 2014-07-31 12:54 -0500
Re: External hard drive usage. NTFS or other file system? Unknown <dog@gmail.com> - 2014-08-14 12:39 +0000
Re: External hard drive usage. NTFS or other file system? Keith Keller <kkeller-usenet@wombat.san-francisco.ca.us> - 2014-08-14 08:15 -0700
Re: External hard drive usage. NTFS or other file system? Tim Watts <tw_usenet@dionic.net> - 2014-07-31 19:08 +0100
Re: External hard drive usage. NTFS or other file system? Robert Heller <heller@deepsoft.com> - 2014-07-31 14:27 -0500
Re: External hard drive usage. NTFS or other file system? Bobbie Sellers <bliss-sf4ever@dslextreme.com> - 2014-07-31 12:43 -0700
csiph-web