From alex at alex.org.uk Sun Aug 1 19:34:39 2004 From: alex at alex.org.uk (Alex Bligh) Date: Sun, 01 Aug 2004 20:34:39 +0100 Subject: Large File Copy to Large ext3 RAID5 Array Often Stalls In-Reply-To: <001301c47746$718bcd20$0b00000a@WIDEBODY> References: <001301c47746$718bcd20$0b00000a@WIDEBODY> Message-ID: --On 31 July 2004 14:36 -0700 Calin Brabandt wrote: > Perhaps this IS a Samba problem. I created a new directory structure on > my RAID5 array from the Linux console and move its contents of large > files to the new subdirectory. Although I need to do more testing, the > problem seem to have vanished--at least when copying new files from my > network Samba clients to the new subdirectory. I am *FAR* from an expert, but I seem to remember that samba isn't particularly efficient at scanning directories, not least because of case-insensitive filename matching. You didn't say what kernel you were using, but if it is 2.4, then ext3 is also not efficient at holding directories with large numbers of files (which is what I think you said you had). It may be that the two problems compound to give you very slow performance. You can try the htree patch for your kernel, then ensure the directory gets indexed - IIRC mv the directory out the way, create a new directory when htree is turned on (which will then be created with the relevant htree index), and mv each of the files in the old directory back into the new directory. If you were using 2.4, but in the intervening period you've upgraded to 2.6 (which has the htree fixes in), I note you've just done effectively that, which may be why the problem has disappeared - just a guess. Alex From calinb at comcast.net Mon Aug 2 08:38:38 2004 From: calinb at comcast.net (Calin Brabandt) Date: Mon, 2 Aug 2004 01:38:38 -0700 Subject: Large File Copy to Large ext3 RAID5 Array Often Stalls In-Reply-To: Message-ID: <000501c4786c$1d6bed70$0b00000a@WIDEBODY> Thanks Alex! I'm running 2.6 from a new install of Fedora Core 2 so it's my understanding that the htree fixes were there all along--unless they came with one of the Core 2 kernel updates. Samba may be responsible for some of my problems, but because I had the same problems with ftp and I couldn't correlate the trouble to anything on the client machines or network, I was ready to blame ext3 or the RAID5. Things seem better now that I recreated the directory structure from the server console and moved the files to the new subdirectories, rather than copying the subdirectories from the Samba clients. Cal -----Original Message----- From: Alex Bligh [mailto:alex at alex.org.uk] Sent: Sunday, August 01, 2004 11:35 AM To: Calin Brabandt; ext3-users at redhat.com Cc: Alex Bligh Subject: RE: Large File Copy to Large ext3 RAID5 Array Often Stalls --On 31 July 2004 14:36 -0700 Calin Brabandt wrote: > Perhaps this IS a Samba problem. I created a new directory structure on > my RAID5 array from the Linux console and move its contents of large > files to the new subdirectory. Although I need to do more testing, the > problem seem to have vanished--at least when copying new files from my > network Samba clients to the new subdirectory. I am *FAR* from an expert, but I seem to remember that samba isn't particularly efficient at scanning directories, not least because of case-insensitive filename matching. You didn't say what kernel you were using, but if it is 2.4, then ext3 is also not efficient at holding directories with large numbers of files (which is what I think you said you had). It may be that the two problems compound to give you very slow performance. You can try the htree patch for your kernel, then ensure the directory gets indexed - IIRC mv the directory out the way, create a new directory when htree is turned on (which will then be created with the relevant htree index), and mv each of the files in the old directory back into the new directory. If you were using 2.4, but in the intervening period you've upgraded to 2.6 (which has the htree fixes in), I note you've just done effectively that, which may be why the problem has disappeared - just a guess. Alex From schmidt at pasemi.com Tue Aug 3 01:40:38 2004 From: schmidt at pasemi.com (Peter Schmidt) Date: Mon, 02 Aug 2004 18:40:38 -0700 Subject: How to determine the size of a existing journal file Message-ID: <6.1.0.6.0.20040802183930.01aae8e8@mail-gw.pasemi.com> I'm wondering if there is a way to figure out what the size of a journal file that was created on an ext3 file system. Any help would be greatly appreciated. Thanks.. peter From zab at zabbo.net Tue Aug 3 03:38:23 2004 From: zab at zabbo.net (Zach Brown) Date: Mon, 02 Aug 2004 20:38:23 -0700 Subject: How to determine the size of a existing journal file In-Reply-To: <6.1.0.6.0.20040802183930.01aae8e8@mail-gw.pasemi.com> References: <6.1.0.6.0.20040802183930.01aae8e8@mail-gw.pasemi.com> Message-ID: <410F08AF.3010605@zabbo.net> Peter Schmidt wrote: > I'm wondering if there is a way to figure out what the size of a journal > file that was created on an ext3 file system. well, one way: # dumpe2fs -h /dev/hda2 | grep 'Journal inode' dumpe2fs 1.34 (25-Jul-2003) Journal inode: 8 # debugfs -R 'stat <8>' /dev/hda2 Inode: 8 Type: regular Mode: 0600 Flags: 0x0 Generation: 0 User: 0 Group: 0 Size: 33554432 [...] 32M, here. - z From simon.oliver at manchester.ac.uk Tue Aug 3 08:26:51 2004 From: simon.oliver at manchester.ac.uk (Simon Oliver) Date: Tue, 3 Aug 2004 09:26:51 +0100 Subject: ext3 performance with hardware RAID5 In-Reply-To: <1089810042.2806.5.camel@laptop.fenrus.com> Message-ID: <008901c47933$a02f3d80$a06e5882@bi.umist.ac.uk> > On Wed, 2004-07-14 at 14:56, Simon Oliver wrote: > > -O dir_index because this is supposed to help file listing > performance > > with Samba shares -R stride=8 because of the 32kb stripe size > > (4kb*8=32kb) > > > > RHEL3 does not include HTREE though (the thing that uses dir_index) > Does any know if the dir_index option will this be included in update 3? Also, any plans for XFS to be included? -- Simon Oliver From susumusa609 at hotmail.com Tue Aug 3 09:38:34 2004 From: susumusa609 at hotmail.com (susu musa) Date: Tue, 03 Aug 2004 09:38:34 +0000 Subject: Recover deleted FILE in EXT-3 FS Message-ID: Hi All Can someone advice me how to recover a deleted file in EXT-3 File system ? Thanks susu _________________________________________________________________ Help STOP SPAM with the new MSN 8 and get 2 months FREE* http://join.msn.com/?page=features/junkmail From Avraham_Shwartz at icomverse.com Tue Aug 3 09:41:28 2004 From: Avraham_Shwartz at icomverse.com (Shwartz Avraham) Date: Tue, 3 Aug 2004 12:41:28 +0300 Subject: Recover deleted FILE in EXT-3 FS Message-ID: <045FA14A4C4F904BA2D5BEDC37E16AF8957F0D@il-tlv-mail01.comverse.com> Hi All Can someone advice me how to recover a deleted file in EXT-3 File system ? Thanks Avraham -------------- next part -------------- An HTML attachment was scrubbed... URL: From sct at redhat.com Mon Aug 9 20:28:38 2004 From: sct at redhat.com (Stephen C. Tweedie) Date: 09 Aug 2004 21:28:38 +0100 Subject: ext3 performance with hardware RAID5 In-Reply-To: <008901c47933$a02f3d80$a06e5882@bi.umist.ac.uk> References: <008901c47933$a02f3d80$a06e5882@bi.umist.ac.uk> Message-ID: <1092083317.2018.221.camel@sisko.scot.redhat.com> Hi, On Tue, 2004-08-03 at 09:26, Simon Oliver wrote: > > On Wed, 2004-07-14 at 14:56, Simon Oliver wrote: > > > -O dir_index because this is supposed to help file listing > > performance > > > with Samba shares -R stride=8 because of the 32kb stripe size > > > (4kb*8=32kb) > > > > > > > RHEL3 does not include HTREE though (the thing that uses dir_index) > > > Does any know if the dir_index option will this be included in update 3? > Also, any plans for XFS to be included? There are no plans to include either of them for update 3. --Stephen From zhuye at pronetway.com Tue Aug 10 01:42:34 2004 From: zhuye at pronetway.com (Zhu Ye) Date: Tue, 10 Aug 2004 09:42:34 +0800 Subject: Conversion / partition from ext2 to ext3 Message-ID: <006701c47e7b$4f5675c0$eb00a8c0@zhuye> Hi, I have installed Red Hat Linux 7.3 with ext2 file system and I have multiple partition. I converted them to ext3 using following command. tune2fs -j -i 0 /dev/hdaX And I modified /etc/fstab as below. LABEL=/ / ext3 defaults 1 1 LABEL=/boot /boot ext3 defaults 1 2 none /dev/pts devpts gid=5,mode=620 0 0 LABEL=/home /home ext3 defaults 1 2 none /proc proc defaults 0 0 none /dev/shm tmpfs defaults 0 0 LABEL=/var /var ext3 defaults 1 2 /dev/hda5 swap swap defaults 0 0 After I rebooted the system, I found / partition still was mounted as ext2. [root at ProEIM root]# cat /proc/mounts rootfs / rootfs rw 0 0 /dev/root / ext2 rw 0 0 # not mounted as ext3 /proc /proc proc rw 0 0 usbdevfs /proc/bus/usb usbdevfs rw 0 0 /dev/hda1 /boot ext3 rw 0 0 none /dev/pts devpts rw 0 0 /dev/hda6 /home ext3 rw 0 0 none /dev/shm tmpfs rw 0 0 /dev/hda2 /var ext3 rw 0 0 [root at ProEIM root]# But it is ext3 when I use df -T. [root at ProEIM root]# df -T Filesystem Type 1k-blocks Used Available Use% Mounted on /dev/hda3 ext3 5036316 335784 4444700 8% / /dev/hda1 ext3 101089 8724 87146 10% /boot /dev/hda6 ext3 66263252 466552 62430684 1% /home none tmpfs 127416 0 127416 0% /dev/shm /dev/hda2 ext3 5036316 101656 4678828 3% /var [root at ProEIM root]# How can I convert / partition from ext2 to ext3? Please help me. Regards, Zhu -------------- next part -------------- An HTML attachment was scrubbed... URL: From adilger at clusterfs.com Tue Aug 10 04:00:47 2004 From: adilger at clusterfs.com (Andreas Dilger) Date: Mon, 9 Aug 2004 22:00:47 -0600 Subject: Conversion / partition from ext2 to ext3 In-Reply-To: <006701c47e7b$4f5675c0$eb00a8c0@zhuye> References: <006701c47e7b$4f5675c0$eb00a8c0@zhuye> Message-ID: <20040810040047.GN18216@schnapps.adilger.int> On Aug 10, 2004 09:42 +0800, Zhu Ye wrote: > After I rebooted the system, I found / partition still was mounted as ext2. > > [root at ProEIM root]# cat /proc/mounts > rootfs / rootfs rw 0 0 > /dev/root / ext2 rw 0 0 # not mounted as ext3 > > But it is ext3 when I use df -T. > > [root at ProEIM root]# df -T > Filesystem Type 1k-blocks Used Available Use% Mounted on > /dev/hda3 ext3 5036316 335784 4444700 8% / > /dev/hda1 ext3 101089 8724 87146 10% /boot > /dev/hda6 ext3 66263252 466552 62430684 1% /home > none tmpfs 127416 0 127416 0% /dev/shm > /dev/hda2 ext3 5036316 101656 4678828 3% /var > [root at ProEIM root]# > > How can I convert / partition from ext2 to ext3? Please help me. Rebuild your initrd (mkinitrd). Cheers, Andreas -- Andreas Dilger http://sourceforge.net/projects/ext2resize/ http://members.shaw.ca/adilger/ http://members.shaw.ca/golinux/ From simon.oliver at umist.ac.uk Tue Aug 10 07:55:04 2004 From: simon.oliver at umist.ac.uk (Simon Oliver) Date: Tue, 10 Aug 2004 08:55:04 +0100 Subject: ext3 performance with hardware RAID5 In-Reply-To: <1092083317.2018.221.camel@sisko.scot.redhat.com> Message-ID: <004301c47eaf$58ee5f10$a06e5882@bi.umist.ac.uk> Stephen C. Tweedie wrote: > > On Tue, 2004-08-03 at 09:26, Simon Oliver wrote: > > > On Wed, 2004-07-14 at 14:56, Simon Oliver wrote: > > > > -O dir_index because this is supposed to help file listing > > > performance > > > > with Samba shares -R stride=8 because of the 32kb stripe size > > > > (4kb*8=32kb) > > > > > > > > > > RHEL3 does not include HTREE though (the thing that uses > dir_index) > > > > > Does any know if the dir_index option will this be included > in update > > 3? Also, any plans for XFS to be included? > > There are no plans to include either of them for update 3. > How can I influence the inclusion plans for RHEL3? -- Simon Oliver From kmshanah at ucwb.org.au Sun Aug 15 04:59:50 2004 From: kmshanah at ucwb.org.au (Kevin Shanahan) Date: Sun, 15 Aug 2004 14:29:50 +0930 Subject: e2fsck hangs while recovering journal Message-ID: <1092545989.12162.24.camel@localhost> Hi, It seems one of my computers died last night and it looks like hard disk trouble. I took the disk out and attached to another system and tried to mount the partition, but the mount doesn't complete and just hangs (process is also unkillable - I had to reboot). So, I tried to run "fsck -fccp /dev/sda2" but the command just hangs after printing "/dev/sda2: recovering journal". Now I'm pretty sure this is because there's bad blocks on the disk, as dmesg shows some disk errors (see attached file). Google tells me status 0x59 means a bad block. I tried to remove the journal to see if the fsck could get any further by using "tune2fs -f -O^has_journal /dev/sda2", but it just told me to recover the journal first (despite using -f). Anyway, I'm kind of stuck. Can anyone give me any clues as to how I can get this filesystem mounted for long enough to try and recover some data? Thanks, Kevin. -------------- next part -------------- scsi1: ERROR on channel 0, id 0, lun 0, CDB: 0x28 00 00 03 2a 3c 00 00 08 00 Current sda: sense = 70 3 ASC=11 ASCQ= 4 Raw sense data:0x70 0x00 0x03 0x00 0x00 0x00 0x00 0x06 0x00 0x00 0x00 0x00 0x11 0x04 end_request: I/O error, dev sda, sector 207420 scsi1: ERROR on channel 0, id 0, lun 0, CDB: 0x28 00 00 03 2a 3d 00 00 07 00 Current sda: sense = 70 3 ASC=11 ASCQ= 4 Raw sense data:0x70 0x00 0x03 0x00 0x00 0x00 0x00 0x06 0x00 0x00 0x00 0x00 0x11 0x04 end_request: I/O error, dev sda, sector 207421 Buffer I/O error on device sda2, logical block 7320 ata2: DMA timeout, stat 0x1 ATA: abnormal status 0x59 on port 0xB807 scsi1: ERROR on channel 0, id 0, lun 0, CDB: 0x28 00 00 03 2a 3e 00 00 06 00 Current sda: sense = 70 3 ASC=11 ASCQ= 4 Raw sense data:0x70 0x00 0x03 0x00 0x00 0x00 0x00 0x06 0x00 0x00 0x00 0x00 0x11 0x04 end_request: I/O error, dev sda, sector 207422 ATA: abnormal status 0x59 on port 0xB807 ATA: abnormal status 0x59 on port 0xB807 ATA: abnormal status 0x59 on port 0xB807 From evilninja at gmx.net Sun Aug 15 10:53:50 2004 From: evilninja at gmx.net (evilninja) Date: Sun, 15 Aug 2004 12:53:50 +0200 Subject: e2fsck hangs while recovering journal In-Reply-To: <1092545989.12162.24.camel@localhost> References: <1092545989.12162.24.camel@localhost> Message-ID: <411F40BE.3030007@gmx.net> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Kevin Shanahan wrote: > Anyway, I'm kind of stuck. Can anyone give me any clues as to how I can > get this filesystem mounted for long enough to try and recover some > data? as always: try to bypass the fs first and get the data to working disk. "dd" or even better "dd_rescue" seem to be the tools of choice here. then, after copying the data to file, you can try "fsck" on this file and see, what it gives. this way you don't have to deal with hardware errors while recovering. Christian. - -- BOFH excuse #449: greenpeace free'd the mallocs -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.5 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFBH0C+C/PVm5+NVoYRAu42AJ9dG2SqqgOKFiho164f9Y6BZ//eQQCfYb7R aQr+izRGzz57yiY6w5x9IqI= =4vYE -----END PGP SIGNATURE----- From kmshanah at ucwb.org.au Mon Aug 16 00:50:27 2004 From: kmshanah at ucwb.org.au (Kevin Shanahan) Date: Mon, 16 Aug 2004 10:20:27 +0930 Subject: e2fsck hangs while recovering journal In-Reply-To: <411F40BE.3030007@gmx.net> References: <1092545989.12162.24.camel@localhost> <411F40BE.3030007@gmx.net> Message-ID: <1092617427.1489.2.camel@localhost> On Sun, 2004-08-15 at 20:23, evilninja wrote: > Kevin Shanahan wrote: > > Anyway, I'm kind of stuck. Can anyone give me any clues as to how I can > > get this filesystem mounted for long enough to try and recover some > > data? > > as always: try to bypass the fs first and get the data to working disk. > "dd" or even better "dd_rescue" seem to be the tools of choice here. > then, after copying the data to file, you can try "fsck" on this file > and see, what it gives. this way you don't have to deal with hardware > errors while recovering. Thanks, this sounds like exactly what I need - apart from more frequent backups 8). I'll try it out when my replacement drive arrives. Cheers, Kevin. From admin at screentiful.it Tue Aug 31 14:33:15 2004 From: admin at screentiful.it (Jois) Date: Tue, 31 Aug 2004 16:33:15 +0200 Subject: Ext3 Errors in laptop with debian sarge Message-ID: <6.0.0.22.0.20040831162352.01b7ce00@pop.e-tree.com> hi, i have a laptop: asus l5800gm, with a dual boot system: windows xp and linux (debian sarge with kernel 2.4.25) These are the errors when I perform fsck.ext3 -y /dev/hdc1. >pass 1: checking inodes, blocks and sizes >error reading block 1090648 (attempt to read block from filesystem resulted >in short read) while reading indirect blocks of inode 360671. Ignore error? >yes >Force rewrite? yes It happens randomly, and sometimes it goes straight through pass 2, 3, 4 and 5; and sometimes it asks me other things as "connect to /lost/found" or something like that, I don't remember right now. It didn't happen last time, yesterday; just that questions "ignore error" and "force rewrite" And, after rebooting, the system date is 2 hours later! I mean, if the current time is 16:00, it says 18:00, both in Windows and in Linux! The computer has been repaired last June, as the mainboard crashed down, and for a month I stopped getting these problems, now they started again, from 21st of August 2004; the Asus customer support told me, when I got the PC repaired, that the hard disk was OK. Could they be wrong? Could I have a damaged hard disk? My situation is particular, as I am a blind user I have to use a special aid, called BRLTTY; unfortunately it doesn't start when the error occurs (give root password for maintenance or type ctrl-d to exit...") and then I type the root password, it says none and there is the prompt with #, so I type brltty and the braille device starts regularly, so I can perform fsck. For this reason, I did not try to mount the partition using a live distribution as Knoppix, because I do not know how to start it with BRLTTY. Could someone please help me with the Ext3 problems? Bye & thanks! Jois