El otro día, por problemas de espacio en mi vetusto Linux, tuve que añadir un segundo disco viejuno que encontré muerto de risa por ahí. Posee la friolera de 30gb de espacio y tras comprobarle feacientemente, ha quedado destinado a ser el disco de almacenamiento temporal de documentos multimedia de gran tonelaje

Por supuesto, el sistema de descargas estaba montado en un directorio, así que s eha eliminado, se ha recreado como punto de montaje del disco y se ha copiado el contenido del viejo directorio en él. Conclusión: el programa sigue funcionando con total normalidad sin enterarse de su nueva ubicación.

Fuentes:

• Particionado, formateado y punto de montaje
  http://www.idevelopment.info/data/Unix/Linux/LINUX_PartitioningandFormattingSecondHardDrive_ext3.shtml

• Cuando hay problemas con el disco:

• Surviving a Linux Filesystem Failures
  http://www.linuxquestions.org/questions/linux-hardware-18/mount-wrong-fs-type-bad-option-bad-superblock-on-devhdc3-373428/
• EXT3: “Bad Superblock”
  http://www.linuxquestions.org/questions/linux-hardware-18/mount-wrong-fs-type-bad-option-bad-superblock-on-devhdc3-373428/

Caros toda as explicações aqui são apenas básicas, eles servem para que voce tenha o conhecimento suficiente para a escolha do melhor sistema de arquivos para o seu computador residencial e também não tenho a pretensão de dizer qual seria a melhor escolha (isso só voce poderá dizer com base em testes)

O que é sistema de arquivos:

Sistema de arquivos é um conjunto de rotinas e estruturas que permite ao sistema operacional ter o controle do acesso aos dados nas unidades de armazenamento.

- Journaling (subst. diário) é a criação de um log com informações sobre os arquivos armazenados no disco como metadata(tamanho, dono e permissões) e as informações dos dados em si, metadadata não é o arquivo propriamente dito é apenas a informação estrutural do mesmo

- EXT3 é a terceira versão do sistema de arquivos EXT que é a mais utilizada hoje em dia, devido a sua maior segurança e ser a evolução do primeiro sistema de arquivos para linux o EXT, ele é considerado o mais seguro (me refiro a integridade dos arquivos) pelo fato do seu sistema de journaling armazenar as informações de metadados e dos dados com isso fazendo com que a recuperação dos dados caso haja um desligamento acidental seja mais eficiente, o sistema de journal do EXT3 possui três modos:

Ordered (padrão) – o journal é atualizado no final de cada operação de escrita, o que diminui o desempenho pelo fado da cabeça fazer a gravação dos dados duas vezes

Wirtebkack – ele armazena apenas informações referente a metadata, o que o tornar o mais rápido dos três modos, mais o torna o mais suscetível a falhas durante um processo de recuperação de arquivos

Journal – é o modo mais seguro e confiável do três (torno a lembrar que é somente a respeito da recuperação dos dados) devido ao fato dele armazenar as informações tando de metadada assim como informações sobre os dados que serão armazenados no disco é o modo que tem o menor desempenho dos três e é o menos utilizado.

- ReiserFS é um sistema de arquivos que tem como característica positiva a utilização de blocos dinâmicos, que nada mais é que to tamanho do bloco aonde o arquivo ira ficar é sempre igual ao seu tamanho (repere no seu computador a propriedade do arquivo e veja que há duas notações uma ao tamanho do arquivo e a outra do tamanho em disco) com o RiserFS isso não ocorre do arquivo ocupar mais espaço em disco do que o necessário, já que o tamanho de cada bloco é ajustado para que o arquivo não ocupe o mais espaço do que o necessário e uma característica negativa (que não pode faltar) é que ele dentre os outros sistemas de arquivos para linux é o que utiliza um maior percentual de processamento, pelo fato dos seus algorítimos serem um pouco mais complexos.

- XFS é um sistema de arquivos de grande desempenho principalmente para arquivos grandes, desenvolvido inicialmente para Unix e adaptado para o linux ele utiliza endereçamento de 64 bits (permite arquivos maiores e aumenta o desempenho do mesmo), possuir journal que grande uma grande velocidade na recuperação dos dados.

- JFS assim como XFS ele é um sistema de arquivos 64 bits ele tem como um característica interessante a capacidade de redimensionar as partições do sistema sem que o computador seja desligado, ele não é muito utilizado pelo fato dele ter um inicio um tanto quanto conturbado devido a grande quantidade de bugs que existia nele, que hoje já foram sanados

- FAT é um sistema de arquivos que foi desenvolvido pela Microsoft para o MS-DOS, que hoje só é utilizado por pen-drivers e outros discos de estado solido (SSD em inglês) como cartões de memória e é virtualmente suportado por todos os sistemas operacionais conhecidos.

- NTFS é o sistema de arquivos mais atual da Microsoft que foi desenvolvido para substituir o FAT devido a falta de segurança e pela sua limitação a reconhecer apenas discos com apenas 2 Gbs, ele assim como os sistemas de arquivos para o linux possui um sistema de recuperação de dados que aumenta as chances de voce recuperar os seus dados e ele teve uma serie de implementações de segurança como o permissões de acesso aos arquivos armazenados no disco.

Amplexos e até o próximo post

PS.: Fiquei tanto tenpo sem posts deivo a falta de moden

Steps:

1. While rebooting the server from the server), use “shutdown -rf now”. This will restrict the system to run an autofsck.

2. Edit the /etc/grub.conf file and place “fastboot” at the end of the kernel line.

3. Edit the /etc/fstab. You can see two numbers at the end of the line for each partition, change the second number to a 0

Example: /dev/sdd2 / ext3 defaults 1 0). This will have the system mount the partition but will not run a check when booting.

En popüler ve kullanıcı dostu Linux dağıtımı olan Ubuntu’yu bilgisayara kurmak için yapılması gerekenler şu şekildedir:

Öncelikle Ubuntu’nun kurulacağı alan ayrılmalıdır. Bunu yapmak için Ubuntu kurulumu esnasındaki Disk bölümleyici uygulama kullanılabildiği gibi benim tavsiyem Windows üzerinden Partition Magic tarzı bir programla bölüm oluşturmanızdır. Bölüm oluştururken yapılacaklardan bahsedeyim biraz da. İlk başta Partition Magic programını veya benzeri bir disk bölümleyici uygulamayı bilgisayara kuruyoruz(Tabi ki bunlar henüz Windows ortamı için). Partition Magic açıldığında bilgisayarda bulunan hard disk(ler)i bölümlerini ve boş alanlarını ayrıntılı şekilde gösterir. Linux için 5-6 GB yer ayırmak yeterlidir. İsteğe bağlı olarak 10 GB da ayrılabilir. Yer ayrılmak istenen hard disk seçilip Partition Magic yardımıyla bu hard disk bölümünün 5 GB lık bölümü ayrılarak yeni biçimlenmemiş bir parça oluşturulur.(Örneğin hard diskim 160 GB ve C sürücüm 60 GB ve D sürücüm 100 GB. D sürücümde 80 GB boş yerim var. Bunun 5 GB kısmını Ubuntu’ya ayırmak istiyorum. Partition Magic yardımıyla D sürücüsünde yeni bir bölüm oluşturulursa Partition Magic eski verileri silmeden 5 GB lık biçimlenmemiş bir bölüm oluşturacaktır.)

Ubuntu kurulması için bölüm oluşturduktan sonra sıra takas alanı denen alanı oluşturmaya geldi. Bu alan yanlış bilmiyorsam eğer bilgisayarın bellek miktarı yetersiz kalırsa hard diskin bir bölümünü ram gibi kullanıp performans sağlamak içindir. Esasında yeni bilgisayarlarda Ubuntu üzerinde çalışırken sistem gözlemcisinden bakınca bu takas alanın hiç kullanıldığını görmedim. Biz yine de ayıralım. Yine 5 GB ayırır gibi 1 GB lık bir alan daha ayırıyoruz istediğimiz hard disk bölümünden. 1 GB takas alanı boyutu sabit bir değer değildir. Genelde 1 GB ayrılır. Kişinin hard disk boyutuna göre farklılık gösterebilir. 512 MB da ayrılabilir. Bu işlemin sonunda 1 GB lık biçimlenmemiş bir alan oluşacaktır. Windows’ta yapılacaklar tamamdır. Artık kuruluma başlanabilir.

1. Ubuntu CD’si  CD/DVD sürücüsüne takılıp bilgisayar yeniden başlatılır.(Tabi eğer ayarlı değilse BIOS ayarlarından bilgisayarın önce CD ile başlatılması sağlanır. Normal başlatılma sırası Floppy – CD – Flash disk – HDD şeklindedir. Ancak bazı bilgisayarlarda farklılık olabilir.)
2. Ubuntu CD’sinden başlatılınca kısa bir beklemenin ardından ekrana dil seçim ekranı gelir. Buradan isteğe bağlı olarak seçim yapılır. Ben Türkçe seçilerek yapılan kurulumu anlatacağım
3. Yine kısa bir beklemenin ardından ekrana gelen yazılar şu şekildedir:

• Bilgisayarınızda herhangi bir değişiklik yapmadan Ubuntu’yu deneyin
• Ubuntu kur
• Diski arıza için kontrol et
• Bellek testi
• Birinci sabit diskten aç

Burada istenirse Ubuntu canlı CD özelliği kullanarak denenebilir. Yani kurulmadan direk CD’den çalışan işletim sistemi olarak. Bunun yararı çökmüş Windows’a format atmadan önce C sürücüsündeki verilere ulaşıp onları yedeklemek…vb dir. Tabi Ubuntu’yu Live olarak çalıştırmak daha fazla bellek tüketimine sebep olur

Bizim seçeceğimiz “Ubuntu kur” seçeneğidir. Buna basınca kısa bir önyükleme görünür. Ardından işletim sistemi dili seçim ekranı gelir.

1. İşletim sistemi dili olarak önerilmiş olan Turkey seçilidir zaten. Forward butonuna basılarak ilerlenir
2. Dünya haritası çıkar ekrana. Buradan da dünya üzerinde bulunduğumuz yere uygun olan bir merkez seçilir. Bu saat ayarı için gereklidir. İstanbul seçilidir zaten. Forward butonuna basılarak ilerlenir.
3. Klavye seçim ekranı karşımıza gelir. Buradan Turkey seçeneğinde 7-8 klavye türü çıkar. Ben Türkçe Q klavye kullanıyorum. Siz kendinize göre seçersiniz. Zaten en yukarıda önerilen seçenekte Türkçe seçilidir. Hiçbir şeye dokunulmadan Forward denilirse klavye ayarlanır.
4. İşte bu nokta kurulumun en önemli ve göz korkutan aslında hiç de öyle olmayan kısmıdır ki ne dedim ben de bilmiyorum:) Burada karşımıza Ubuntu’nun kurulacağı alan seçimi çıkar. Bu alanı kolayca geçebilmemizin anahtarı Windows’tan ayarladığımız alanlardır.

Burada kesinlikle yapılmaması gereken “Bütün diske kur” seçeneğini seçmektir. En alttaki seçenek olan “Bölümleri elle belirt(gelişmiş)” seçeneğinin seçilmesi en sağlam yöntemdir.

1. “Bölümleri elle belirt(gelişmiş)” seçilip “Forward”  butonuna basıldığında hard disk bölümlerimiz boyutları, adları ve boş alanları ile birlikte ekrana gelir. Kendi C, D, E sürücülerimizin yanında yeni ayırdığımız 5 GB ve 1 GB lık bölümler de boş alan olarak görünür. Öncelikle 1 GB lık kısım seçilir ve alttan “Yeni disk bölümü” butonuna basılır. Ekrana bir pencere açılır. “Megabayt cinsinden bölüm boyutu” kısmına dokunmadan(1 GB takas alanı ayrıldıysa 1024 civarı bir değer olacaktır) “Nasıl kullanılacağı” kısmına gelinir ve oradan “takas alanı” seçilir. Sizin Ubuntu dağıtımı versiyonunuza göre burada “swap” yazıyor da olabilir.  “Ok” butonuna basılır ve tekrar hard disk ekranına dönülür. Yalnız bu kez bir yenilenme olur sayfada ve 1 GB lık kısmın yanında “swap” yazar. Bunun ardından sıra Ubuntu’nun kurulacağı hard disk parçasına belirlemeye gelir. 5 GB lık kısmı zaten bunun için ayırmıştık bu nedenle onu seçip üstteki gibi “Yeni disk bölümü” butonuna basılır. Az önceki takas alanı yaparken çıkan pencerenin aynısı açılacaktır. Burada yine “Megabayt cinsinden bölüm boyutu” kısmına dokunmadan “Nasıl kullanılacağı” kısmına gelinir ve “Ext3 günlüklü dosya sistemi” veya varsa “Ext4 journaling file system” seçilir.(Bu Ubuntu sürümünüze bağlıdır. Ext4 dosya sistemi yeni çıkmıştır ve Ext3′e göre daha hızlı olduğu söylenmektedir. Seçim sizin tabi)

Nasıl kullanılacağı kısmı ayarlandıktan sonra onun altındaki “Bağlama noktası” kısmına gelinir ve elle / yazılır veya listeye basılarak oradan / (back slash)karakteri seçilir. “Ok” butonuyla bu işlem tamamlanıp tekrar hard disk ekranına dönüldüğünde 5 GB lık kısmın yanında bir tik işareti görülür. İşte buraya Ubuntu kurulacaktır. “Forward” denerek ilerlenir.

1. Adınızı ister ve bilgisayar şifresi falan ister. Bunlar ileride kullanılacaktır. Unutmayacağınız bir şeyler koymanızı tavsiye ederim. Hele şifre her daim lazım olacak.
2. “Forward” butonuna basıldığında Microsoft Windows XP yazan bir ekran gelir. Burada hiçbir şeyin değiştirmeden devam edilir. İçe aktarma ekranıdır bu. Ama pek kullanılmaz.(Ben hiç kullanmadım)
3. “Forward” denerek ilerlenir ve kurulum başlar. Bu noktadan kurulum bitimine kadar tek yapmanız gereken beklemektir.
4. Kurulum bitince “Kurulum tamamlandı” yazar ve sistemi yeniden başlatma isteğini gösterir size. Siz de yeniden başlatırsınız ve artık Ubuntu kullanıma hazırdır.
5. Yeniden başlatınca Windows başlangıcının aksine GRUB denen Ubuntu’nun kendi başlatma yöneticisinin ekranı gelir. Burada bilgisayarda yüklü olan işletim sistemlerinin listesi vardır
6. Bu ekranda en üstte yeni kurduğumuz Ubuntu “Other Operating Systems” kısmında da “Microsoft Windows XP” bulunur. Tabi bu en basit haliyle. Kişinin bilgisayarındaki işletim sistemleri daha fazla olabilir. Ona göre seçim yapılır. Ubuntu seçildiğinde kullanıcı adı ve şifre ister. Bunlar kurulumdan önce belirlediğimiz ad ve şifredir. Girildiğinde gnome masaüstü tüm ihtişamıyla karşımıza çıkar.

]]> http://chuletasalvapor.wordpress.com/2009/09/23/pasa-de-ext3-a-ext4-sin-formatear/ Wed, 23 Sep 2009 02:31:39 +0000 dejotaplayerz http://chuletasalvapor.wordpress.com/2009/09/23/pasa-de-ext3-a-ext4-sin-formatear/

Partisi adalah sebuah bagian dari memori atau media penyimpanan yang terpisah secara logis yang berfungsi seolah-olah bagian tersebut terpisah secara fisik. Media penyimpanan yang dapat dipartisi adalah memori (baik itu memori fisik ataupun memori maya oleh manajer memori sistem operasi), hard disk, magneto-optical disk (MO Disk), dan beberapa flash memory. Meskipun demikian, istilah partisi saat ini digunakan untuk merujuk pada bagian dari hard disk.

Partisi yang pertama saya kenal yaitu FAT, partisi yang umum dipakai pada Sistem Operasi Windows, Setelah Partisi FAT saya mengenal Partisi FAT 32, dan NTFS pada beberapa tahun belakangan. Untuk Operating System (OS) Windows, partisi yang berlaku untuk sitem yang lebih tinggi tidak bisa dipakai untuk sistem operasi yang berada dibawahnya. contoh, untuk Windows 95 digunakan Partisi Dengan Format FAT16, tidak bisa memakai Partisi dengan format NTFS yang digunakan untuk windows XP.

Setelah masuk kuliah, ternyata saya baru menyadari akan banyaknya jenis partisi pada hardisk. ketika itu saya mencoba menginstall Vector Linux. klau tidak salah ada sekitar 70-an lebih option partisi yang disediakan oleh partition tool yang dipakai oleh vector linux. Itu sebagai contoh kasus saja.

Semua sistem operasi mulai dari DOS, Windows, Macintosh dan turunan UNIX memiliki Sistem berkas sendiri untuk meletakkan file dalam sebuah struktur hirarki. Contoh dari sistem berkas termasuk di dalamnya FAT, NTFS, HFS dan HFS+, ext2, ext3, ISO 9660, ODS-5, dan UDF. Beberapa sistem berkas antara lain juga journaling file system atau versioning file system.

Sistem berkas juga menentukan konvensi penamaan berkas dan peletakan berkas pada stuktur direktori.

Setiap sistem operasi dan sistem berkas memiliki sebutan tersendiri untuk menyebut partisi. Sebagai contoh, MS-DOS menggunakan istilah partition, sementara keluarga Windows NT menggunakan istilah volume. Hal ini disebabkan oleh Windows NT yang memiliki kemampuan untuk membentuk satu volume yang terdiri dari beberapa partisi terpisah, daripada sistem operasi MS-DOS yang hanya dapat membuat satu volume untuk satu partisi.

sumber: http://id.wikipedia.org/wiki/Sistem_berkas

Terminamos a homologação do sistema de arquivos EXT4 para uso em ambiente corporativo, tanto em desktops como em servidores com grandes volume de I/O de disco. (servidores de bancos de dados por exemplo)

Esse foi o primeiro resultado positivo de nosso esforço para colaborar com desenvolvimento e homologação do Linux Ubuntu em ambiente corporativo. (anuncio de 01/07/2009)

Por conta disso, já entregamos soluções que utilizam o EXT4 para nossos clientes, e estamos fomentando a troca gradativa do EXT3 para EXT4 tanto em nossos data centeres e escritórios, mas também em clientes com ambiente legados.

Os testes envolveram um trabalho de de 1.000 horas em laboratório nosso e em at wild, em clientes que se dispuseram a participar da maratona de homologação.

Mais dados, inclusive de performance serão divulgados ainda esta semana.

Postingan kali ini adalah sekedar tutorial sederhana dari gw cara melakukan automount partisi windows di Linux baik NTFS maupun FAT32 yang umum dipakai pada saat startup. Jadi begitu start linux kita ga perlu lagi melakukan mount secara berulang-ulang,mending kalo Cuma satu partisi nah kalo banyak????bakal repot pastinya. Sudah gtu bakal ga praktis juga kalo kita punya lagu2 kesayangan di partisi windows ini dan harus selalu dimasukkan ke playlist audio player secara manual  di linux setiap kali kita masuk ke linux. SANGAT TIDAK PRAKTIS BUKAN?ok langsung ke tips sederhana dari gw.

Tutorial ini menggunakan distro ubuntu tapi umumnya juga bisa diterapkan di distro lain karena kita akan melakukan “konfigurasi” pada file fstab.

1. Seperti biasa buka konsole dan login sebagai root su, bisa juga dengan sudo tapi saya lebih suka su sih
2. Buka file fstab (di ubuntu terletak di /etc/fstab) dengan text editor seperti gedit atau vi atau apapun terserah anda. Misal gedit /etc/fstab. Isinya kurang lebih seperti ini:

# /etc/fstab: static file system information.
> #
> # <file system> <mount point> <type> <options> <dump> <pass>
> proc /proc proc defaults 0 0
> # /dev/sda1
> UUID=ec08eda5-1ac7-402f-84e7-77c0c26e3206 / ext3
> relatime,errors=remount-ro 0 1
> # /dev/sda3
> UUID=8817-CDBD /dos vfat utf8,umask=007,gid=46 0 1
> # /dev/sda4
> UUID=E291-3013 /windows vfat utf8,umask=007,gid=46 0 1
> # /dev/sda2
> UUID=1899d45b-8e12-4ef5-9b04-b1b248c7f8c1 none swap
> sw 0 0
> /dev/scd0 /media/cdrom0 udf,iso9660 user,noauto,exec,utf8 0 0

Coba perhatikan diatas!ada keterangan /windows ada /dos lalu ada vfat,umask=007,gid,ext3 dsb. Tidak usah bingung!yang jadi perhatian kita adalah UUID=xxxxxxxxxxxx kemudian type partisinya yaitu (VFAT untuk FAT dan ntfs untuk NTFS) serta perintah umask dan gid. Saya rada lupa sih tapi kalo ga salah umask=007 adalah “perintah” tertingginya alias ini memerintahkan partisi tersebut agar bisa langsung di mount saat startup dan juga bisa ditulisi seperti partisi linux saja.Saya mengasumsikan anda sudah paham maksud dari /dev/sda(x) dll itu.

3. Pasti bingung,darimana kita tau UUID kita?karena masing2 tipe partisi UUIDnya pasti beda lho, caranya kita bisa tau dengan mengetikan blkid di konsole(harus dalam root)
4. Tahap berikutnya(bisa dengan buka tab konsole baru),login sbg root lagi dan buat folder baru sebagai tujuan mount partisi misalnya diletakkan di /media/windows jadi kita bikin folder namanya windows di dalam folder /media/ . CARANYA ketik mkdir /media/windows di konsole.
5. Berikutnya adalah kita harus tahu partisi mana yang MAU kita automount kan,partisi disini “biasanya” dikenal sebagai sda(x) ataupun sdb(x) untuk yang punya 2 buah harddisk. X berati nomor mulai dari satu sesuai jumlah partisi. Kita gunakan perintah fdisk -l di konsole lagi. Nanti bakal muncul informasi yang kurang lebih seperti ini:

Disk /dev/sda: 120.0 GB, 120034123776 bytes
255 heads, 63 sectors/track, 14593 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0×80d2f3ee

Device Boot Start End Blocks Id System
/dev/sda1 * 1 2833 22756041 83 Linux
/dev/sda2 2834 3165 2666790 82 Linux swap / Solaris
/dev/sda3 3166 6934 30274492+ b W95 FAT32
/dev/sda4 6935 14593 61520917+ b W95 FAT32

Berdasarkan keterangan diatas terlihat bahwa ukuran harddisk adalah 120GB,jumlah partisinya ada 4 yaitu sda1-sda4 dan type file yang non linux hanya ada FAT32,NTFS pun bisa di mount juga kok dan juga bisa support baca tulis seperti FAT32 Cuma saratnya harus sudah terinstall ntfs-3g di linuxnya. Untuk kernel 2.6.27 keatas(Ubuntu 8.04 keatas) sepertinya sudah default terinstal.

6. Sekarang kita bisa lihat partisi (nomor) keberapa yang akan di automount,contoh diatas adalah sda3 dengan type FAT32. Sesuai dengan tahap2 di poin sebelumnya dimana kita sudah tau nomor UUID(poin 3),kita bisa menulis manual di file /etc/fstab (poin 2.) dan juga “folder” tujuan untuk atumount (poin 4).
kurang lebih ketik seperti ini di fstab:

# /dev/sda3
> UUID=8817-CDBD /media/windows vfat utf8,umask=007,gid=46  0     1

Tinggal diurut aja bacanya,diliat dari keterangan diatasnya dari
# <file system> <mount point> <type> <options> <dump> <pass>

Alamat “file” yang di mount terletak di /dev/sda3, UUID yang didapat lewat blkid yaitu 8817-CDBD, typenya vfat. Berikutnya yang rada ribet adalah keterangan utf8,umask=007,gid=46 0 1.
7. Kalau tipenya nadalah NTFS maka ganti saja vfat dengan ntfs

artinya:
umask=007 will set the permissions on the partition to 770:
owner = read/write/execute
group = read/write/execute
others = no permissions

umask=007,gid=46 – allows read/write/execute permissions for the users from the plugdev group.

8. Apalagi neh muncul plugdev????gw jg ga tau sih…tapi yang jelas karena kita mengetikkan keterangan gid=46 dan umask itu dimana partisi yang kita mount memilik hak baca tulis oleh “user yang terdaftar” di plugdev so mau ga mau kita HARUS “mendaftarkan” user account kita di plugdev, caranya

Addgroup username plugdev, username bisa diganti nama username anda.
9. Done,sekarang anda bisa memount secara otomatis partisi windows anda di linux saat startup selain itu juga support baca tulis sehingga ga da beda seperti jalan di windows asli.

CMIIW

There are several how-tos out there on how to convert an ext3 file system to ext4. Those all use tune2fs to convert the file system in place, adding the features exclusive to ext4. While this is certainly the quickest approach, it doesn’t give you all the performance of a fresh installation with ext4, since the extends feature will only be used for new files. To get the most out of ext4, you should completly recreate the partition. I’ll show you how to do so savely:

Preparation
To make sure the data isn’t changing while we’re copying it, we boot from a Jaunty live CD. You’ll need to use the current version, since older versions of Ubuntu do not support ext4.
Burn the CD and boot from it.
You need to backup the data on your ext3 partition while preserving all the meta data (ownership, permissions etc). The best tool for this is rsync. Let’s assume your ext3 partition is /dev/sda1 and your backup partition is /dev/sdb1. The backup partition should use ext3 too, otherwise some meta data might be lost. Do not use FAT!

Backing up your data
Mount the devices (if they aren’t already)

sudo mkdir /media/ext3
sudo mkdir /media/backup
sudo mount -t ext3 /dev/sda1 /media/ext3/
sudo mount -t ext3 /dev/sdb1 /media/backup/

Create the backup

sudo rsync -axvh --progress /media/ext3/ /media/backup/

Reformat
This is the easiest part.

sudo umount /media/ext3/
sudo mke2fs -t ext4 /dev/sda1

Restoring your data from backup
Mount the new file system

sudo mkdir /media/ext4
sudo mount -t ext4 /dev/sda1 /media/ext4/

Restore the data

sudo rsync -axvh --progress /media/backup /media/ext4/

Since the backup isn’t needed anymore, unmount it.

sudo umount /media/backup/

Reinstall grub
If the partition you converted contains the /boot directory, you also need to reinstall grub.

sudo grub-install --root-directory=/media/ext4/ \
/dev/sda1
sudo grub
root (hd0,0)
setup (hd0)

Here, hd0,0 is the device/partition number that corresponds to sda1. The most reliable way to find out which partition number you need is the find command. Try
find /boot/grub/menu.lst
in grub.

Change fstab

• Open /etc/fstab as root.
• Change the entry for your partition from ext3 to ext4.

Change menu.lst
Those changes are only needed if the partition you converted is your root partition.

• Find out the new UUID.
  
ls -l /dev/disk/by-uuid | grep sda1


• Open /boot/grub/menu.lst as root.
• Change the UUID in the root= kernel option for every entry.

I did this on Ubuntu 9.04 (Jaunty). Did it work for you, too? Or did it mess up your harddrive and you want to kill me? Tell me in the comments!

Una cosa que me toca hacer cada vez que reinstalo Ubuntu (que poquito me gusta hacerlo, aunque es menos pesado que tener que hacer lo mismo con Windows) es tener que añadir la partición que uso de almacenamiento para que se monte con el inicio del sistema, y es algo que nunca me acuerdo como se hace.

Es bien sencillo, en 2 pasos lo tenemos listo:

- Con la unidad montada, ejecutamos en consola el comando ‘df‘ que nos devolverá cierta información sobre las unidades montadas que tenemos. Deberemos localizar la que nos interesa ya sea por el nombre o la capacidad en su defecto y después ver donde está el sistema de ficheros, en mi caso /dev/sda2.

- Mediante la orden ‘sudo gedit /etc/fstab‘ editamos dicho fichero añadiendo una linea con la siguiente estructura:

donde_se_encuentra     donde_montamos      filesystem      defaults     0      0

En mi caso la linea a añadir es la siguiente:

/dev/sda2    /media/data    ext3    defaults   0   0

Fácil y sencillo para toda la familia.

Enabling quota for users may be useful to limit the use of space disk.
This works on ext3 and ext4 filesystems. With journaled quota, you do not need to issue a quotacheck command even if your server has been unproperly halted.

Install the appropriate Debian packages :

root@localhost:~# apt-get install quota quotatool

Once installed, please modify your /etc/fstab to enable journaled quota on your selected partition :
Add the following line to the appropriate partition :

usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0

Once added, please run the following commands :

root@localhost:~# touch /aquota.user /aquota.group
root@localhost:~# chmod 600 /aquota.*
root@localhost:~# mount -o remount /

If you remount your filesystem, it prevents you from rebooting the server.

Please verify that journaled quota is enabled :

root@localhost:~# quotaon -vug /your_partition
root@localhost:~# quotacheck -vugm /your_partition

A menudo las herramientas de particionado a nuestra disposición nos resultan incomodas, limitadas o complicadas de usar, todos los sistemas operativos incluyen herramientas de particionado, pero raras veces son tan completas como quisiéramos, en esas ocasiones puede sernos de utilidad una herramienta como GParted.

Para realizar el proceso detallado en este tutorial necesitaremos el Cd Live GParted, si vuestro ordenador permite arrancar desde USB podréis ahorraros un CD.

1) Como es costumbre con los Live CD, introducimos el CD y configuramos en la BIOS el arranque desde el CD-ROM, guardamos los cambios y salimos de la BIOS.

2)Este Live Cd nos presenta una interface bastante amigable y cómoda de usar, nada complicada, seleccionamos la opción “GParted” para iniciar el programa y acto seguido nos aparecerá la tabla de particiones, en función de si el disco está vacío o si disponemos de particiones ya creadas el aspecto que presentará sera distinto pero fácil de entender.

1. La opción “New” nos permitirá crear nuevas particiones, mientras que la opción “delete” nos permitirá borrarlas convirtiéndolas en espacio no asignado, la opción redimensionar nos permitirá reducir o aumentar el tamaño de una o mas particiones así como mover de sitio la partición si deseamos colocarla al final o delante de otra.
2. La tabla de particiones se nos muestra con un aspecto similar a un gráfico, pudiendo observar las distintas particiones, su posición, el espacio que ocupan o su nombre.
3. El nombre de la partición si tiene.
4. El sistema de archivos de la partición, “extended” hace referencia a la memoria Swap.
5. El tamaño total de la partición, su espacio ocupado y su espacio libre.

Si deseamos crear una partición, deberemos situarnos sobre el espacio del disco duro no asignado (en gris) y o bien con botón derecho o bien con el icono “New” podremos configurar sus características.

Podremos ajustar su tamaño tanto introduciendo el numero exacto en la casilla “Tamaño nuevo (MB)” como manualmente mediante el gráfico superior (algo mas impreciso que introduciendo el numero), la opción “Crear como” hace referencia a si deseamos una partición primario o lógica, en “Sistema de archivos” seleccionaremos el que se ajuste a nuestras necesidades, Ext2 o Ext 3 si ha de albergar un Linux, Swap si ha de ser la de intercambio.
Si deseamos ponerle una etiqueta a nuestra partición para tenerla diferenciada podemos hacerlo escribiendo aquellos que nos parezca acertado, el “espacio libre precedente“ y “el espacio libre a continuación” lo dejamos a 0.

Pulsamos en “add” y ya tendremos configurada la partición, podremos crear aquellas que necesitemos, en mi caso, dos para la instalación de una distribución Linux, por lo que necesitaré otra partición Swap.

Una vez creadas las particiones necesarias, procederemos a guardar los cambios, pulsaremos en “apply” para que se apliquen los cambios y escriba las particiones en la tabla de particiones, cuando termine se habrán realizado correctamente todos los cambios realizados y podremos salir del CD-Live con la opción “exit“.

Karmic Koala is out and it looks great! The latest version of one of the more popular Linux distributions has some major changes. The ‘What’s new in Karmic’ series will look at the most important ones and see what they mean.

Ext4 Filesystem

Ext3, which was the default Linux filesystem for many years now, has given way to the faster and more efficient Ext4. Ext4 has been in development for almost 3 years now, and, while it was added as an option in Ubuntu 9.04 (Jaunty Jackalope), it now comes packaged along with the latest edition as the default filesystem.

The biggest difference between the two filesystems is the volume size. Ext3’s 32-bit block numbers and 4KB data blocks give us a volume limit of 16TB. While the block size remains the same in Ext4, the block numbers are 48 bit, which gives us a whopping 1024PB or 1EB (1 Exabyte).

Ext4 also reduces fragmentation in files and allocates multiple blocks for a single file, by choosing contiguous blocks on the disk. This is possible due to the new extents scheme which replaces the existing block mapping scheme. Hence large files perform better.

Unallocated blocks are skipped during diskchecks and this improves speed. Pre-allocation and delayed allocation of blocks also enhance performance. Delayed allocation as the name suggests is delaying block allocation until all data is going to be written to disk. A slight disadvantage of this is if the system crashes before allocation, in which case there is a loss of data. But home users need not worry about this. 

There are a lot of other complicated details which i will not go into. Ultimately, the Ext4 filesystem is much better, faster, larger, more efficient, more reliable, and very stable. If you have an ext3 filesystem or lower do not hesitate to go ahead and upgrade to Ext4. You will see a marked improvement in the performance of your machine. If you are using Ubuntu, a fresh install of Karmic Koala does the job for you automatically. To do it manually, here are the steps.

First make a note of the filesystem ID you would like to upgrade to ext4. You can find this by going to System -> Partition Editor in Jaunty. A list of partitions will show up on your screen. It should look like this-

sda1 is my id. Type the following in a terminal -

sudo tune2fs -O extents,uninit_bg,dir_index /dev/sda1

sudo fsck -pf /dev/sda1

sudo mount -t ext4 /dev/sda1 /mnt

Then open the fstab file by typing

gksu gedit /mnt/etc/fstab

Change ext3 to ext4 and then save and exit.

Now in the terminal type,

sudo grub-install /dev/sda

Remember, just  sda, no numbers. If all goes well then your system will be ready. Reboot and enjoy!

I worked 20 hours straight to prepare an old Gateway tower with 280 gig drive for an Ubuntu 9.10 install. The Gateway was hopelessly unbootable in Windows because of some operating system problem (not hardware problem). I realized that I should drop down to C: and type format, to just reformat the entire drive which took about 30 minutes. I forgot that I should have typed FDISK first, to set up the necessary partitions, but it was too late.

I downloaded an ISO image for GNOME partition and burned it to a bootable CD.

I booted the GNOME partition utility. It took me literally hours to guess at what I should do.

Finally I realized that Ubuntu expects THREE partitions. The first is referred to as ROOT but the actual name you give it in the partition process is simpl “/” (forward slash omitting the double quotes).

The second partition will have NO name, but you will designate its format TYPE as swapfile, and that will become the system swap file.
The third partition is called /home.

I had no idea how much space to allocate, so I gave 512 for root and 512 for swap, and the huge remainder for /home.

The software warned me that root and swap were too small. Just for the heck of it, I changed root and swap to 5000, and it worked.

Then it took me dozens of tries installing Ubuntu. I had a 2008 Ubuntu disc and a 9.10 disc I had created last week. Time after time, I would boot from each one, say INSTALL UBUNTU, and it would chug away for a long time, and then die with no messages. FINALLY, but why or how I dont know, the 9.10 unstall brought up an Ubuntu desktop with what looked like all kinds of error messages. There were TWO icons on the desktop which I had never seen before. One said EXAMPLES which I never looked at. I was about to give up in dispair when I decided to click on the icon marked Ubuntu. AS IT TURNS OUT, that is the icon which COMPLETES the install process, and it first brings you into its OWN partition software. I cant exactly remember what I did next. I do remember that GNOME partition utility offered me literally 20 DIFFERENT format types to choose from for each partition (one of which is swap-file). I had no clue what to choose. I did some google searching on my other machine (and you NEED a windows machine attached to the internet to look these things up). I discovered that format type ext3 is good for Debian Linux (and Ubuntu is a flavor of Debian) so I went with ext3.

Finally, I had a working bootable pure Ubuntu 9.10 machine.

My next Sisyphean agony task was to connect it to the Internet. The tower already had an Ethernet card. I was informed that IF I connected a cable from the ethernet to the router, then Ubuntu would automatically sense the ethernet connection and configure it. The ethernet board DID light up and pulse when I connected the cable to my Westell router. But Ubuntu never recognized it. Later I read that only certain ethernet cards will be automatically recognized, so I guess the one in my tower was not a compatible one. My next choice was to try and get my Belkin USB WiFi adapter working. I went to ADMINISTRATION -> SYNAPTIC PACKAGE MANAGER and keyed ndiswrapper into the search. I was prompted to place the Ubuntu install cd into my drive. It kept giving me errors. I placed my Belkin adapter install cd in the drive and looked around for the ini driver file, but could not see it. Finally, out of despair, I plugged the Belkin wifi adapter into the USB and LIKE MAGIC Ubuntu recognized it, asked me to choose a connection and enter the WEP key, which I have now done many times with other machines and during the Wubi Ubuntu install on Windows. So now I had a working Ubuntu machine connected to the Internet.

I went to J&R computer store up the street and found a $10 USB 4 gig memory stick by http://www.dane-elec.com which is REALLY cheap. I recently discovered that the Sancor memory sticks that I love actually have a lot of SOFTWARE on them, which only works under windows, AND if you have a password set on the device, then Ubuntu wont even read it.

I had to go into synaptic manager and search on usb for some installs. Again, it was trial and error, and I dont know what finally made it work, but finally, Ubuntu recognized the memory stick.

TODAY, I had to meet my step-son near at Borders Book Store next to Penn Station. This was lucky for me, because while I waited for him, I browsed the computer book section on the second floor. I found a $9 small Linux handbook reference of commands by Daniel J. Barrett (O Reilly publishers) so I got that to practice all the hundreds of Linux commands that I dont know but need to know.

I browse various books on Ubuntu and Mysql.

I found the key thing I needed to know for my next project, to install LAMP (Linux Apache MySQL PhP) on my new Ubuntu desktop.

All one needs to do is enter TERMINAL, and key in

sudo tasksel

(you will be prompted for your system password next)

A task selection window will pop up with many packages available for install, one of which is LAMP. Now here is a mystery that took me an hour to solve. You can arrow down to highlight LAMP, but you dont have a clue in the world as to how to CHECK it for install. Google revealed the secret: YOU SIMPLY PRESS THE SPACE BAR AND IT IS MARKED CHECKED FOR INSTALL. Then you press ENTER and the install proceeds, Apache, MySQL and PhP. You will be prompted for the MySQL admin password.

Next I wanted to install phpmyadmin to administrate mysql, creating databases, tables, etc. I went to SYSTEM-> Synaptic Manager and searched on phpmyadmin. I think there is only one. BUT HERE IS A SECRET. Once your LAMP install finishes, then power down and boot back up; otherwise, certain files will remain locked by processes and the phpmyadmin install will fail.

Once phpmyadmin installs (and you will again be prompted for that pesky password), then you can open firefox and key in http://localhost/phpmyadmin

You will be PROMPTED FOR A USER NAME AND PASSWORD. This is tricky and confusing. You have to key in the user name “root” without quotes, and then give the password that you have been using all along throughout this. Now you are in phpmyadmin, and you can go to Privileges, and create a new user and give him all possible privileges.

Then, I found a $50 book, also by O’Reilly, entitled
“Head First – PHP & MySQL”. THIS book looks like the perfect way to teach myself how to develop php mysql applications which will have proper security (which is a very complex topic).

I didnt purchase that book tonight, because I wanted to come home and see first if I could successfully install LAMP and phpmyadmin, which I just now finished.

The Head First book gave me links to try the book on line free for 45 days. Money is kind of tight, so I decided to see how far I can go without any additional purchases, but if I get things working, I will buy the book one day. It is interesting that O’Reilly also offers a link to an on line university which can grant degrees in computer science.

I signed up at Oreilly for a trial access of the electronic version of the above-mentioned book.

Here is something I tried which crashed my Ubuntu. I wanted to copy the introduction to the book, from the OReilly site, and PASTE it into an Open Office Document. The first time Open Office died, and the file was shown as requiring recovery. I did the recovery, opened it again with Open Office Document, and then the whole Ubuntu system crashed. I rebooted, and had to go through a long complicated reconstruct procedure. I thought perhaps I would have to install everything from scratch, but fortunately, Ubuntu repaired itself. Moral to story is paste into GEDIT and save, which I guess eliminates whatever hidden characters are in the webpage which cause Open Office to crash the system.

There, I found a link to download all the source code from the book.

I downloaded the zipped file of all the source code.

Next, I went to the source code for chapter 1. They have a beginning version (which I guess you make changes to, and then a final version).

I tried to copy and paste report.html to var/www which is where Apache wants to find pages loaded from local host. I was not allowed to paste because I do not have superuser privileges. So, I went to terminal and typed SUDO GEDIT, which gets me into gedit with super powers, so I can now navigate to the unzipped code in my DOCUMENTS in a special folder I created, open report.html, then do a save as to var/www.

THIS WORKED, because now, when I open Firefox and key in localhost/report , it runs the Chapter 1 example webpage.

FOR MY NEXT TRICK, I must figure out how to copy a chapter 1 tutorial image of a dog (fang.jpg) from the location where it unzipped (in a folder in my Documents) and copy it into var/www where I do not have Admin privileges. SO, I open my OReilly handbook of Linux commands and find the CP command. I open a TERMINAL session and key in SUDO CP which I know will executed it with super admin privileges, but NOW I do not have a clue as to what the PATH names are for the file source and destination.

I am permitted to copy fang.jpg from the unzipped file into DOCUMENTS, just to simplify things.

Using PLACES -> COMPUTER, which brings up the FILE BROWSER, I first navigate to home/Documents and I RIGHT CLICK on the fang.jpg image, and click on PROPERTIES which will SHOW ME that the path is
/home/bryan/Documents/fang.jpg (I use bryan as my user name, but that is my step-son’s name). Next I do the same right click trick on any file which is already in var/www such as report.html, and I discover THAT path, which tells me that the command I should issue in TERMINAL is

sudo cp /home/bryan/Documents/fang.jpg var/www/fang.jpg

BUT IT HANGS… NOTHING HAPPENS.

Then I realize that I should close my other applications, because they are somehow LOCKING UP something that the CP command needs. Sure enough, once I close a bunch of applications, the SUDO CP command asks me for the PASSWORD and then copies fang.jpg to var/www

I can now bring up the image by opening Firefox and typing

localhost/fang.jpg

into the browser address field.

NOW that fang.jpg is in var/www, if I once again open Firefox and type
localhost/report.html

the exercise 1 page loads and also shows the picture of the little dog fang.jpg

I still need to copy report.php and style.css from the place where they unzipped to var/www. Since I can launch gedit with sudo and have superadmin powers, I may open them in that fashion and do a save as.

I ran the finished example, hoping that it would actually send me an email.

I realized that I must examine my php.ini and somehow reconfigure it. I went to some pains to locate that file, but did not post notes as I went, and now I must reconstruct what I did. I still have not succeeded in getting the app to actually send an email through php. But I feel it is more essential for me to move on to other lessons in the HEAD FIRST book.

How long does it take to mkfs a 10GB disk with all the different filesystems out there?

See my test results here using the new guestfish sparse / filesystem support. btrfs is “best” and ext3 comes off “worst”.

As a test this is interesting, but it’s not that relevant for most users — they will be most interested in how well the filesystem performs for their workload, which is not affected by mkfs time and hard to measure in general benchmarks anyway.

Update

In response to Stephen’s comment, I retested this using a memory-backed block device so there is no question about whether the host backing store affects the test:

$ for fs in ext2 ext3 ext4 xfs jfs reiserfs nilfs2 ntfs msdos btrfs hfs hfsplus gfs gfs2
    do guestfish sparse /dev/shm/test.img 10G : run : echo $fs : sfdiskM /dev/sda , : \
        time mkfs $fs /dev/sda1
    done
ext2
elapsed time: 1.45 seconds
ext3
elapsed time: 2.71 seconds
ext4
elapsed time: 2.58 seconds
xfs
elapsed time: 0.13 seconds
jfs
elapsed time: 0.27 seconds
reiserfs
elapsed time: 0.33 seconds
nilfs2
elapsed time: 0.08 seconds
ntfs
elapsed time: 2.07 seconds
msdos
elapsed time: 0.14 seconds
btrfs
elapsed time: 0.07 seconds
hfs
elapsed time: 0.17 seconds
hfsplus
elapsed time: 0.17 seconds
gfs
elapsed time: 0.84 seconds
gfs2
elapsed time: 2.76 seconds

When I was at University, a philosophy lecturer remarked rather sagely that University is the last place people can go to learn for the sake of learning.

That’s sort of correct, but not always so. People can fumble through their jobs on a day to day basis learning what they have to, but they can also work along the basis of trying to soak up as much information as they can along the way. I’m not always a knowledge sponge – particularly if my caffeine quota is on the light side for the day, but I like to think I learn the odd thing here and there.

In the spirit of knowledge acquisition, here’s a few smaller things I’ve learned recently:

• When simulating network connectivity problems, there’s a big difference between yanking the network cable and shutting down the network interface. (I was doing the interface shutdown, another person was doing the network cable unplug – and our results didn’t correlate.) Lesson: When escalating a case to vendor support, always spell out how you’re simulating the “comms failure” a customer is having.
• The ‘bigasm’ utility starts to fall in a heap and becomes extremely unreliable once you exceed about 2100 GB of data generated for a single file. Lesson: When setting out to generate 2.3+ TB of backup data, create a bunch of files and have a bigasm directive to generate a smaller amount of data per file.
• When setting up tests that will take a couple of days to run, always triple check what you’re about to do before you start it. Lesson: If you make a typo of 250 files at 100 GB each instead of 250 files at 10 GB each, bigasm/NetWorker won’t interpolate what you really meant.
• There’s a hell of a difference between Solaris 10 AMD release 2 and release 8. Lesson: If wanting to get a Solaris 10 AMD 64-bit OS working in Parallels Desktop for Mac v5 with networking, go for release 8. It will save many forehead bruises.
• ext3 is about as “modern” a filesystem as I am an elite sportsperson. Lesson: If wanting to achieve decent operational activities with backup to disk under Linux, use XFS instead of ext3.
• All eSATA is not created equal. Lesson: When using an motherboard SATA -> eSATA converter, make sure the dual drive dock you order doesn’t work as a port multiplier.

Which filesystems have the largest metadata overhead[1]? We can find out using guestfish sparse file support and a short shell script:

#!/bin/sh -

guestfish=fish/guestfish
testimg=/mnt/tmp/test/test.img
size=1G

for fstype in ext2 ext3 ext4 xfs ntfs msdos ; do
    $guestfish <<EOF
      sparse $testimg $size
      run
      sfdiskM /dev/sda ,
      mkfs $fstype /dev/sda1
EOF
    echo -n "$fstype: "
    du $testimg
done

The results:

Notes

[1] Yes, there are shortcomings in the methodology. What this really measures is how many blocks are written by the mkfs program. Even if you believe the figures, this only measures the initial overhead, but you can use the same technique to measure the overhead of storing (eg) lots of small files, or whatever is appropriate for your workload.
[2] Note the difference between ext2 and ext3/4 seems to be entirely down to the size of the journal, which is a kind of metadata overhead, but one that you can easily control.
[3] Needs three small patches to libguestfs to get these filesystems to work.
[4] This seems off the scale — needs further investigation.

Maybe you recently discovered that just after a fresh install of the last version of your favorite Linux Distro, apps like explore2fs, Acronis True Image or the Ext2/3 Driver for Windows couldn’t read no more your Ext3 Filesystem, Acronis True Image even complaining about “Partition Errors”.

This is due to the fact that most of new distros (starting from version 8.10 concerning Ubuntu) are using the inode 256 structure by default for their filesystem. There’s pretty much nothing to do about that, the only way to switch back to the well-supported inode 128 is to reformat your hard drive…

Acronis True Image Home 2009 (build 9709) is the only version that supports inode 256 (Business/Workstation 2009 don’t, nor 2010 versions), you can still switch to CloneZilla or Partimage which are excellent (and free) alternatives to True Image or Ghost.

Recently when I made an exasperated posting about lengthy ext3 check times and looking forward to btrfs, Siobhán Ellis pointed out that there was already a filesystem available for Linux that met a lot of my needs – particularly in the backup space, where I’m after:

• Being able to create large filesystems that don’t take exorbitantly long to check
• Being able to avoid checks on abrupt system resets
• Speeding up the removal of files when staging completes or large backups abort

That filesystem of course is XFS.

I’ve recently spent some time shuffling data around and presenting XFS filesystems to my Linux lab servers in place of ext3, and I’ll fully admit that I’m horribly embarrassed I hadn’t thought to try this out earlier. If anything, I’m stuck looking for the right superlative to describe the changes.

Case in point – I was (and indeed still am) doing some testing where I need to generate >2.5TB of backup data from a Windows 32-bit client for a single saveset. As you can imagine, not only does this take a while to generate, but it also takes a while to clear from disk. I had got about 400 GB into the saveset the first time I was testing and realised I’d made a mistake with the setup so I needed to stop and start again. On an ext3 filesystem, it took more than 10 minutes after cancelling the backup before the saveset had been fully deleted. It may have taken longer – I gave up waiting at that point, went to another terminal to do something else and lost track of how long it actually took.

It was around that point that I recalled having XFS recommended to me for testing purposes, so I downloaded the extra packages required to use XFS within CentOS and reformatting the ~3TB filesystem to XFS.

The next test that I ran aborted due to a (!!!) comms error 1.8TB through the backup. Guess how long it took to clear the space? No, seriously, guess – because I couldn’t log onto the test server fast enough to actually see the space clearing. The backup aborted, and the space was suddenly back again. That’s a 1.8TB file deleted in seconds.

That’s the way a filesystem should work.

I’ve since done some (in VMs) nasty power-cycle mid-operation tests and the XFS filesystems come back up practically instantaneously – no extended check sessions that make you want to cry in frustration.

If you’re backing up to disk on Linux, you’d be mad to use anything other than XFS as your filesystem. Quite frankly, I’m kicking myself that I didn’t do this years ago.