[opensuse-factory] About ext3 as default and periodic fs checks
The periodic ext3 fs checks at boot are driving me nuts. I know they can be disabled. Couldn't they be performed at shutdown instead of boot? Whenever someone starts the computer it means 100% they need it right then. Having ext3 perform fs checking on a 300 GB full drive is nothing that any user will tolerate easily. OTOH whenever someone shuts down the computer, there's a 95% chance that it doesn't need it right then anymore (it's not a reboot). However, there is a small problem with my suggestion. There is a chance that the fs check at shutdown finds an error that needs user intervention. In that case I think there must be a way to mark the volume dirty and go forward with the shutdown. Then at boot do again a fs check and let the user decide. These would be the steps, from boot: 0. boot 1. is fs dirty? then fsck and prompt user how to fix errors 2. run OS 3. at shutdown: should the fs be checked (verify mount count or period) if yes then fsck with "automatic fix" option. Did the fsck find errors that cannot be fixed automatically? If yes mark fs dirty. Shutdown. This way, the user would never have to wait at boot, and the filesystem would still be checked periodically. What do you think? --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Am Montag, 30. Oktober 2006 16:41 schrieb Silviu Marin-Caea:
Whenever someone starts the computer it means 100% they need it right then.
As we learned from past experiences: if someone stops the computer it means they want it in 90% cases off :) Greetings, Stephan --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On Mon, Oct 30, 2006 at 05:41:29PM +0200, Silviu Marin-Caea wrote:
The periodic ext3 fs checks at boot are driving me nuts. I know they can be disabled.
Couldn't they be performed at shutdown instead of boot? [...] These would be the steps, from boot:
0. boot 1. is fs dirty? then fsck and prompt user how to fix errors [...] This way, the user would never have to wait at boot, and the filesystem would still be checked periodically.
No.
What do you think?
I would prefer the following: If a fsck at boot time occurs - kill the splash screen (usually the computer boots in 3 minutes, now it's still unchanged after 10 minutes. Something must be broken. Poweroff/on. Does still not boot. Damn Linux, doesn't work. Changing OS because Linux doesn't work for me). - present an option "you can interrupt the current fsck by pressing ESC. This means the fsck is repeated upon next boot." - As user, I know what will happen next time I boot and prepare for it. If you trust hardware and file systems, you might want to present an easy option to the user to disable those checks alltogether. I wouldn't recommend that. Rasmus --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Silviu, On Monday 30 October 2006 07:41, Silviu Marin-Caea wrote:
The periodic ext3 fs checks at boot are driving me nuts. I know they can be disabled.
What is the check period? How is it measured? In reboot cycles? Calender time elapsed since previous check? Some combination of these? Something else?
Couldn't they be performed at shutdown instead of boot?
Whenever someone starts the computer it means 100% they need it right then.
Having ext3 perform fs checking on a 300 GB full drive is nothing that any user will tolerate easily.
OTOH whenever someone shuts down the computer, there's a 95% chance that it doesn't need it right then anymore (it's not a reboot).
The problem I see with that is that when a system starts up, it's in as stable and pristine a state as it will ever be. On the other hand, when shutting down, it's distinctly more likely that there will be something amiss, even in the kernel, and a file system check in such a state could do more harm than good.
...
Randall Schulz --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On Mon, Oct 30, 2006 at 07:59:07AM -0800, Randall R Schulz wrote:
On Monday 30 October 2006 07:41, Silviu Marin-Caea wrote:
The periodic ext3 fs checks at boot are driving me nuts. I know they can be disabled.
What is the check period? How is it measured? In reboot cycles? Calender time elapsed since previous check? Some combination of these? Something else?
tune2fs -l <device> Look for "Maximum mount count" and "Check interval". Rasmus --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On Monday 30 October 2006 17:53, Rasmus Plewe wrote:
If a fsck at boot time occurs - kill the splash screen (usually the computer boots in 3 minutes, now it's still unchanged after 10 minutes. Something must be broken. Poweroff/on. Does still not boot. Damn Linux, doesn't work. Changing OS because Linux doesn't work for me).
This is exactly what will happen for almost any random average user. They will not press [Esc] for more information. They will press reset. --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Silviu Marin-Caea wrote:
The periodic ext3 fs checks at boot are driving me nuts. I know they can be disabled.
Couldn't they be performed at shutdown instead of boot?
Whenever someone starts the computer it means 100% they need it right then.
Having ext3 perform fs checking on a 300 GB full drive is nothing that any user will tolerate easily.
You could theoretically automate those checks in the background. But you would need the fs in LVM and free space in your VG. 1. Make a snapshot of your fs 2. check the snapshot fs (you may have to replay the logs first (mount-umount)) 3. check return code for possible errors 4. delete snapshot 5. if no error returned, reset the 'last-checked' flags of the original fs (with tune2fs). --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On Tuesday 31 October 2006 9:36 pm, Silviu Marin-Caea wrote:
On Monday 30 October 2006 17:53, Rasmus Plewe wrote:
If a fsck at boot time occurs - kill the splash screen (usually the computer boots in 3 minutes, now it's still unchanged after 10 minutes. Something must be broken. Poweroff/on. Does still not boot. Damn Linux, doesn't work. Changing OS because Linux doesn't work for me).
This is frighteningly true, here are 2 suggestions, we probably need them both. 1) While an fsck is occurring in non-verbose startup mode we need a message to that effect with a progress bar. 2) On shutdown; if a routine fsck WOULD happen next reboot, the OS asks if the user minds converting the shutdown to a reboot. The fsck can happen on a fresh boot but not when the computer is needed. Once booted the system puts up a window a la Mac, waits 2 minutes and shuts back down automatically. With the fsck done, the next reboot happens normally. How long does it have to take to boot Linux? Solaris used to usually boot quickly, unless you touched /reconfigure, then it would go through all the hardware detection. Is anyone working on an "express boot" that gets the system up quickly based on stored hardware configs? With all the hotplug functionality we now have hardware detection at boot is a bit unnecessary. michaelj PS: I'm sad about loosing reiserfs -- Michael James michael.james@csiro.au System Administrator voice: 02 6246 5040 CSIRO Bioinformatics Facility fax: 02 6246 5166 No matter how much you pay for software, you always get less than you hoped. Unless you pay nothing, then you get more. --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
PS: I'm sad about loosing reiserfs
...especially if that means, if I interpret this thread correctly, the "just-in-case" fschecks are back :-(((((((((((. Is this really the case? IMHO this is a giant step back into the past. Not even Windows does that. Alex --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Michael James
On Tuesday 31 October 2006 9:36 pm, Silviu Marin-Caea wrote:
On Monday 30 October 2006 17:53, Rasmus Plewe wrote:
If a fsck at boot time occurs - kill the splash screen (usually the computer boots in 3 minutes, now it's still unchanged after 10 minutes. Something must be broken. Poweroff/on. Does still not boot. Damn Linux, doesn't work. Changing OS because Linux doesn't work for me).
This is frighteningly true, here are 2 suggestions, we probably need them both.
1) While an fsck is occurring in non-verbose startup mode we need a message to that effect with a progress bar.
2) On shutdown; if a routine fsck WOULD happen next reboot, the OS asks if the user minds converting the shutdown to a reboot. The fsck can happen on a fresh boot but not when the computer is needed. Once booted the system puts up a window a la Mac, waits 2 minutes and shuts back down automatically. With the fsck done, the next reboot happens normally.
Could you add those to the feature wishlist in our wiki so that it does not get forgotten?
How long does it have to take to boot Linux? Solaris used to usually boot quickly, unless you touched /reconfigure, then it would go through all the hardware detection. Is anyone working on an "express boot" that gets the system up quickly based on stored hardware configs?
With all the hotplug functionality we now have hardware detection at boot is a bit unnecessary.
michaelj
PS: I'm sad about loosing reiserfs
You're not loosing it - just the default is changed. Feel free to change it for your installs, Andreas -- Andreas Jaeger, aj@suse.de, http://www.suse.de/~aj/ SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
I'm glad to see Michael James think the same way as me. About two days ago I've migrated from Suse 10.1 to Windows because Suse 10.1 takes to boot 4 minutes (not joking!) and Windows Xp about 1 minut (not joking!). I think one of the worst things about Suse is the TIME! The same happens when you run YAST. It is too slow. It is improductive. I hoppe next releases have "boot express". I tried to put it in a wish list but I couldn't; I don't understand how to send it to a wish list. Sorry about it. Hernan Lorenzo F. galicia360@gmail.com
Alex wrote:
PS: I'm sad about loosing reiserfs
...especially if that means, if I interpret this thread correctly, the "just-in-case" fschecks are back :-(((((((((((.
Is this really the case? IMHO this is a giant step back into the past. Not even Windows does that.
Alex --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
That was the main reason I went to reiserfs from ext2, fsck.ext2 of a 20G drive was painful back then, 10x20G, I'd need to power up just before going to bed. Regards Sid. -- Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support Specialist, Cricket Coach Microsoft Windows Free Zone - Linux used for all Computing Tasks --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
That was the main reason I went to reiserfs from ext2, fsck.ext2 of a 20G drive was painful back then, 10x20G, I'd need to power up just before going to bed.
Precisely. Ext2 was and is pretty much useless on large filesystems because one needs to take a day off work every 20 boots. With reiserfs, at boot only the journal gets replayed (automatically when mounting). A complete reiserfsck takes a lot longer too. Nobody deemed it necessary to run a reiserfsck at boot. Why is it deemed necessary to force a run of e3fsck at boot? It's a journalling fs, it's expected to cope. Volker -- Volker Kuhlmann is list0570 with the domain in header http://volker.dnsalias.net/ Please do not CC list postings to me. --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
This is frighteningly true, here are 2 suggestions, we probably need them both. 1) While an fsck is occurring in non-verbose startup mode we need a message to that effect with a progress bar. 2) On shutdown; if a routine fsck WOULD happen next reboot, the OS asks if the user minds converting the shutdown to a reboot. The fsck can happen on a fresh boot but not when the computer is needed. Once booted the system puts up a window a la Mac, waits 2 minutes and shuts back down automatically. With the fsck done, the next reboot happens normally. Could you add those to the feature wishlist in our wiki so that it does not get forgotten?
I don't want to sound too trolling ;), but this is just a workaround to a very unsatisfactory situation, nervertheless catapulting Linux back into archaic computing times for the average desktop user. I'm not a file system specialist and I don't want to become one, but why on earth has a file system in the 21st century to perform time consuming checks - with absolutely no other reasons for this behaviour apart from the fact that the system has been booted x times? If there is reason to assume that something might be wrong, then let the fs check itself - but not just because the computer has been sut off cleanly off a couple of times. As I said, I am no specialist, and I am no ReiserFS evangelic, but for me one of the biggest advantages of ReiserFS was that I was rid of the absolutely maddening periodic fs checks. While I understand the rationale for changing the default fs from a maintainer's point of view, I consider it major flaw from the end-users pov. Alex --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On Fri, Nov 03, 2006 at 09:07:13PM +0100, Alex wrote:
I'm not a file system specialist and I don't want to become one, but why on earth has a file system in the 21st century to perform time consuming checks - with absolutely no other reasons for this behaviour apart from the fact that the system has been booted x times? If there is reason to assume
There is no _need_ to do so. It is just an extra level of safety to do regular checks to detect inconsistencies as early as possible. This is like doing backups: You don't _have_ to do them but it is generally considered a wise choice to do so.
that something might be wrong, then let the fs check itself - but not just because the computer has been sut off cleanly off a couple of times.
The problem is that often you have to check the filesystem to _find_ that there is a reason that something might be wrong. Obviously the other solution is to wait until a inconsistency has managed to garble all your data.
As I said, I am no specialist, and I am no ReiserFS evangelic, but for me one of the biggest advantages of ReiserFS was that I was rid of the absolutely maddening periodic fs checks. While I understand the rationale for changing the default fs from a maintainer's point of view, I consider it major flaw from the end-users pov.
This is not a problem of the filesystem. There are the same advantages and disadvantages to check a ReiserFS filesystem as there are for ext3 filesystems or any other comparable one. It's just that distributions tend to handle them differently but actually there is no reason not to change that if the weights of advantages and disadvantages are rethought. Robert -- Robert Schiele Dipl.-Wirtsch.informatiker mailto:rschiele@gmail.com "Quidquid latine dictum sit, altum sonatur."
Hi, On Fri, 3 Nov 2006, Robert Schiele wrote:
On Fri, Nov 03, 2006 at 09:07:13PM +0100, Alex wrote:
I'm not a file system specialist and I don't want to become one, but why on earth has a file system in the 21st century to perform time consuming checks - with absolutely no other reasons for this behaviour apart from the fact that the system has been booted x times? If there is reason to assume
There is no _need_ to do so. It is just an extra level of safety to do regular checks to detect inconsistencies as early as possible. This is like doing backups: You don't _have_ to do them but it is generally considered a wise choice to do so.
that something might be wrong, then let the fs check itself - but not just because the computer has been sut off cleanly off a couple of times.
The problem is that often you have to check the filesystem to _find_ that there is a reason that something might be wrong. Obviously the other solution is to wait until a inconsistency has managed to garble all your data.
As I said, I am no specialist, and I am no ReiserFS evangelic, but for me one of the biggest advantages of ReiserFS was that I was rid of the absolutely maddening periodic fs checks. While I understand the rationale for changing the default fs from a maintainer's point of view, I consider it major flaw from the end-users pov.
This is not a problem of the filesystem. There are the same advantages and disadvantages to check a ReiserFS filesystem as there are for ext3 filesystems or any other comparable one. It's just that distributions tend to handle them differently but actually there is no reason not to change that if the weights of advantages and disadvantages are rethought.
... and not to forget: you can always abort an unwanted "regular" ext3 filesystem check during boot with CTRL-C. Cheers -e -- Eberhard Moenkeberg (emoenke@gwdg.de, em@kki.org) --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
that something might be wrong, then let the fs check itself - but not just because the computer has been sut off cleanly off a couple of times. The problem is that often you have to check the filesystem to _find_ that
There is no _need_ to do so. It is just an extra level of safety to do regular checks to detect inconsistencies as early as possible. This is like doing backups: You don't _have_ to do them but it is generally considered a wise choice to do so. there is a reason that something might be wrong. Obviously the other solution is to wait until a inconsistency has managed to garble all your data. ... and not to forget: you can always abort an unwanted "regular" ext3 filesystem check during boot with CTRL-C.
Well, for me the question remains why most other filesystems seemingly don't need this regular checks and are nevertheless considered safe enough?? It's obviously not my decision and I will simply stay with ReiserFS, but I would consider it a bad decision to let loose an filesystem on the average home user which forces him to do regular checks. (Situation might be differnt for servers which run 24/7 and where absolute data integrity is a must.) If I shut down the computer I want it shut down (and safe to disconnect from power), if I turn it on, I want to have the desktop running as fast as possible. I - as a home user - don't want to supervise the boot- or shutdown-process, whether the computer decides to do some time consuming self-diagnosis. Just my opinion, based on the fact that Ext2 self checks drove me nuts in former days ;). Alex --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Saturday 2006-11-04 at 05:57 +0100, Alex wrote:
Well, for me the question remains why most other filesystems seemingly don't need this regular checks and are nevertheless considered safe enough??
It's obviously not my decision and I will simply stay with ReiserFS, but I
Wrong, reiserfs is also checked on boot; it just runs faster. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) Comment: Made with pgp4pine 1.76 iD8DBQFFTLigtTMYHG2NR9URAkbxAJ98U5yjFDDz+Q8Egm7ydGKu0XD+6QCeNLDu 3rWAr+CfxqkNGPuNZrPs5kw= =nTho -----END PGP SIGNATURE----- --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Monday 2006-10-30 at 17:41 +0200, Silviu Marin-Caea wrote:
Couldn't they be performed at shutdown instead of boot?
Whenever someone starts the computer it means 100% they need it right then.
When I halt my computer I'm waiting in a hurry to be off; when I'm booting I'm not going anywhere. Plus, when I start working, I want it to be in full order. It is fine the way it is, though faster would be nice :-) - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) Comment: Made with pgp4pine 1.76 iD8DBQFFTLi0tTMYHG2NR9URAssYAJwPmg6+iF59eoc3v8qfXDODn9qSsQCdEM3h +wyVndJA2TkyVDUJuR8Ajho= =awJD -----END PGP SIGNATURE----- --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Friday 2006-11-03 at 22:59 +1300, Volker Kuhlmann wrote:
With reiserfs, at boot only the journal gets replayed (automatically when mounting). A complete reiserfsck takes a lot longer too. Nobody deemed it necessary to run a reiserfsck at boot.
Why is it deemed necessary to force a run of e3fsck at boot? It's a journalling fs, it's expected to cope.
If I understand the script correctly, all partition types are checked at boot, even reiserfs. It's just that it runs so fast you don't notice. Furthermore, there is a type of corruption in reiserfs that goes undetected during that check, that needs and external fsck run. It causes denied permission on files even for root. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) Comment: Made with pgp4pine 1.76 iD8DBQFFTmNitTMYHG2NR9URApJ7AJ9jWyQ/ZrpMy+BoGXDyWmNAVCIIKQCeOtT9 2UyzeVNA0uIAuA+r4G3M+1E= =8sBz -----END PGP SIGNATURE----- --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Carlos E. R. a écrit :
Why is it deemed necessary to force a run of e3fsck at boot? It's a journalling fs, it's expected to cope.
good question.
If I understand the script correctly, all partition types are checked at boot, even reiserfs. It's just that it runs so fast you don't notice.
nope. reiser check only the journal, as etx3 do also probably each start
Furthermore, there is a type of corruption in reiserfs that goes undetected during that check, that needs and external fsck run. It causes denied permission on files even for root.
any file system can have such problem if a failure happen at the bad moment. but it should be easy to have a big red windows saying "warning, at the next reboot a complete chas to be done and can be long" so one can make a boot dedicated to this task jdd -- http://www.dodin.net http://dodin.org/mediawiki/index.php/GPS_Lowrance_GO --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
JDD, On Sunday 05 November 2006 23:49, jdd wrote:
Carlos E. R. a écrit :
Why is it deemed necessary to force a run of e3fsck at boot? It's a journalling fs, it's expected to cope.
good question.
The journal is just a redundant log of operations performed on the file system. It takes an fsck to verify the file system's integrity and replay the journal entries needed to repair any damage found. As an aside, for reasons I don't understand, my 10.2 box sometimes has journal entries to replay upon restart (sometimes quite a few of them) even when it did not crash. I don't know if that's a sign of a problem or normal for Reiser file systems.
...
jdd
Randall Schulz --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Randall R Schulz schreef:
JDD,
On Sunday 05 November 2006 23:49, jdd wrote:
Carlos E. R. a écrit :
Why is it deemed necessary to force a run of e3fsck at boot? It's a journalling fs, it's expected to cope. good question.
The journal is just a redundant log of operations performed on the file system. It takes an fsck to verify the file system's integrity and replay the journal entries needed to repair any damage found.
As an aside, for reasons I don't understand, my 10.2 box sometimes has journal entries to replay upon restart (sometimes quite a few of them) even when it did not crash. I don't know if that's a sign of a problem or normal for Reiser file systems.
I have seen this also happen several times... (no serious effects noticed about this..)
...
jdd
Randall Schulz --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iD8DBQFFT08JX5/X5X6LpDgRAm4FAKClaeBxiUbyyN07UwJlD2FY7zbpDACgmwzg REBllWF+Bdr3D6uNWXD+eqo= =3GB8 -----END PGP SIGNATURE----- --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Monday 2006-11-06 at 08:49 +0100, jdd wrote:
Carlos E. R. a écrit :
If I understand the script correctly, all partition types are checked at boot, even reiserfs. It's just that it runs so fast you don't notice.
nope. reiser check only the journal, as etx3 do also probably each start
If you check the "/etc/init.d/boot.rootfsck" script, it does a fsck run regardless of partition type (and if I'm not mistaken on every boot). It's up to fsck to decide what to do then. And, when the system was not properly closed, it runs fsck -f, for "force".
Furthermore, there is a type of corruption in reiserfs that goes undetected during that check, that needs and external fsck run. It causes denied permission on files even for root.
any file system can have such problem if a failure happen at the bad moment.
Yes, but the difference for reiserfs is that the fast test during boot does not notice any problem, and there is.
but it should be easy to have a big red windows saying "warning, at the next reboot a complete chas to be done and can be long" so one can make a boot dedicated to this task
I don't know how easy is it to know... each partition may have a different day for the check run. Don't you have to access the partition internal structures to know when the next "deep" check has to be done? - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) Comment: Made with pgp4pine 1.76 iD8DBQFFT9batTMYHG2NR9URAltlAJwPriou47AmVX0xzy3xYZxtQuOhUgCfVf2H bTlTv5qI428vpkijr+YDTaw= =F7XR -----END PGP SIGNATURE-----
participants (16)
-
Alex
-
Andreas Jaeger
-
Carlos E. R.
-
Eberhard Moenkeberg
-
Heiko Helmle
-
Hernán Lorenzo Fernández
-
jdd
-
Michael James
-
Randall R Schulz
-
Rasmus Plewe
-
Robby (M9.) Verberne
-
Robert Schiele
-
Sid Boyce
-
Silviu Marin-Caea
-
Stephan Kulow
-
Volker Kuhlmann