On Tuesday 15 March 2005 04:52 am, Darryl Gregorash wrote:

**Never** fsck a mounted partition.

Doh! <slaps head> Of course I know that, Darryl, and my apologies to Kastos. This embarrassing little instinct "bites me" in the "arse" *every single time*! Here's how it goes: a) I mount the filesystem b) I fsck (or variant) it c) I see the following: WARNING!!! Running _<fsck et al>_ on a mounted filesystem may cause SEVERE filesystem damage. Do you really want to continue (y/n)? d) I slap my head in a "Doh!" moment and type "n" for "no". e) I umount the filesystem f) I proceed, vowing to never do that again. g) lather, rinse, repeat - Carl -- _______________________________________________________________________ C. E. Hartung Business Development & Support Services http://www.cehartung.com/ carlh@cehartung.com Dover Foxcroft, Maine, USA Public Keys 68396713 & F8207216 Reg. Linux User #350527 http://counter.li.org/ ----->>>>>http://www.fsf.org/campaigns/supportlinuxbios.html <<<<<-----

Damon Register wrote:

Darryl Gregorash wrote:

...

**Never** fsck a mounted partition.

Ok, I'll bite. Why not? Please don't think that I wish to argue, I just want to learn.

Damon Register

Well then let it put me in these words: If you are to clean a house and (dozens or how many ever) people use it, how can you ever get it clean? Actually you can't, right? Same with a filesystem in use! Cheers Martin

On Tue, 2005-03-15 at 09:39, Damon Register wrote:

Darryl Gregorash wrote:

...

**Never** fsck a mounted partition. Ok, I'll bite. Why not? Please don't think that I wish to argue, I just want to learn.

I think either your clock or timezone are off by three hours. Even if you check the tread on http://lists.suse.com/archive/suse-linux-e/2005-Mar/ and look at the timestamps you will see a difference. You email is three hours into the future. This can cause a problem with some spam checking software. I know by looking at the logs for spamassassin it will show points for it. This is the entry from my log for your email: Mar 15 06:38:48 pc1 amavis[15910]: (15910-08) spam_scan: hits=2.904 tests=DATE_IN_FUTURE_03_06. As you can see it added almost 3 points and may flag it as spam on some systems depending on how tight they have theirs set. Mine is set to flag at 3 points and reject at 5. This is one reason why it important to have your clock set correctly. -- Ken Schneider UNIX since 1989, linux since 1994, SuSE since 1998 * Only reply to the list please* "The day Microsoft makes something that doesn't suck is probably the day they start making vacuum cleaners." -Ernst Jan Plugge

On Tue, 2005-03-15 at 12:32 -0500, Damon Register wrote:

You were right on both counts. I didn't catch it at first but I looked at my timezone and found it was set to the wrong (Pacific) timezone. I also discovered the bios clock was off too. With both corrected, could you please tell me if this is correct now?

Damon Register

I have been away from my PC for awhile but it appears to be fine now. -- Ken Schneider

The Tuesday 2005-03-15 at 08:52 -0800, Randall R Schulz wrote:

On Tuesday 15 March 2005 06:39, Damon Register wrote:

...

Darryl Gregorash wrote:

...

**Never** fsck a mounted partition.

Ok, I'll bite. Why not? Please don't think that I wish to argue, I just want to learn.

The kernel has much cached file system information in RAM. If fsck changes anything on the disk then, the data the kernel holds no longer matches that on the disk, yet the kernel does not "know" this and will access the device as if the information it holds accurately reflects the state, structure and contents of the drive. Bad consequences are very likely to ensue.

That's true. However, a theoretical point raises to my mind. O:-) It should be possible for the kernel to know what is happening and update its internal buffers. In other words, the kernel could keep track of what fsck is doing, update or flush or empty buffers as appropriate, and conversely, be able to write to disk whatever the rest of the OS seems fit, and informing fsck of what has changed. It should be possible, albeit very complex. And slow. But I have seen more complex feats... like, on a double computer, updating half of a mirror disk with the computer continuing normal processing, then suddenly switching to the other, updated half, without interrupting service. Roughly so, but almost correct explanation. -- Cheers, Carlos Robinson

The Wednesday 2005-03-16 at 18:42 -0800, Randall R Schulz wrote:

It is probably possible in principle, though probably not worth it in practice.

Probably. It would make sense for high reliance 24*7 systems, perhaps.

There are two issues that I can think of:

1) Programs like fsck (more precisely, one of the file system-specific fscks) access disks in so-called "raw" mode, bypassing the kernel buffer pool. This makes them rather faster and they don't have to worry that their writes are deferred until the buffer cache must be flushed.

Correct.

2) Much of the kernel's internal state that is derived from disk contents is interpreted. I.e., the disk contents are read and rather than being used directly are interpreted and used to build internal data structure. Even if the kernel took note of every write that fsck performed a write, it is not necessarily easy for it to reflect all those changes internally. For one thing, fsck may use several writes to effect what is a single logical change to the file system. But the kernel just sees a bunch of independent writes. You'd need a separate protocol by which fsck said something like <superbBlockUpdate> write() ... write() ... write() ... </superBlockUpdate> or <freeListRebuild> write() ... write() ... write() ... </superBlockUpdate>.

Yes, fsck would have to work trough the kernel, not directly. Or design a protocol, as you say.

Once you design the structures and protocols that permit this, you're probably going to just throw up your hands, declare it not worth the effort and declare: Thou shalt not repair mounted file systems.

X'-) -- Cheers, Carlos Robinson

Hi all! Whooo! Now that seemed to be a huge thread I kicked off! Too bad I was swaeting over my suse-box with no e-mail trying to save everything. Well it looks like I made it! Thanks to all your tips! Well, to start off, reiserfsck seemed to have made things worse. I was not able to boot at all. Safe settings would hang at some point as would normal boot. I then tried rescue on the DVD and while it (naturally) started up the thing would not see the disk. Some rescue, I thought. Then I grabbed my Knoppix (3.4) and that was not able to see the disks either (It might be that Knoppix uses an older LVM version, but I didn't dig into that). Then I noticed that the thing would boot when I gave option 1 (into init 1) and would see the disks ok. However I could not access the network. As I wanted to take a backup anyway though I got me a spanking new 160 GB disk and installed a clean 9.1. After installing LVM packages the new system was able to mount and read the disks ok! The rest is of course hours of copying. Neato, all my .kde settings are working too! Including my bunch of kmail filters to keep the mailing lists in order... There was one interesting thing though: after I inserted the new (SATA) disk I could not boot beyond BIOS. I had to (after a colleague tipped me off) put the two LVM (IDE) disks on different chains - rather than the same as they were before - in order to boot correctly. Mind you that before the crash the disks were running on the same chain for a year or so. A quick answer to the "heat" thread. The computer as a whole was definitely not overheated as I dismantled the side panels a long time ago. However, I had put the HDs one above and one below the floppy, so maybe that aged them or something. BTW, How do I now check the old disks if they are ok? Any suggestions for tools? So now that I had a new system I wanted to upgrade to 9.2 (as it was a fresh 9.1 install anyway) but I get stuck in yast. but that's definitely for another thread... Cheers and a big thanks again folks, Kostas

On Tuesday 22 March 2005 20:05, Darryl Gregorash wrote:

Konstantinos Georgokitsos wrote:

...

<snip> or something. BTW, How do I now check the old disks if they are ok? Any suggestions for tools?

<snip>

badblocks, contained in the e2fsprogs package, can do an exhaustive non-destructive test of the entire surface of your drives. See also man e2fsck. I am not aware of any similar utils for other file systems.

Try the smartctl command from the smartmontools package. PaulH -- Paul Hewlett (Linux #359543) Email:`echo az.oc.evitcaten@ttelweh | rev` Tel: +27 21 852 8812 Cel: +27 72 719 2725 Fax: +27 86 672 0563 --

On Tuesday 15 March 2005 05:50 am, Martin Deppe wrote:

my guess as of you telling your problem appears after about half an hour is that you might have cooling problems which occur now after taking a while to show up. Possible? If yes you should definitely see to cool your discs down. I mean I had similar problems for a while and I am aware of them now.

Good luck Martin

Look at all your fans and dont forget the video card! It may have a very small fan on it. Be sure and clean all the heatsinks whether fanned or not. You can check out the heat problem by booting with something like Knoppix and playing with your system to see if it dies as it warms up. No harddrive is involved. Another repair option is to boot from cd1 or dvd, play like you're going to do an install then when the window pops up select the option 'Repair the system'. At next screen i would do the auto repair and let Yast to most of the work for you. Then like the others suggest, backup your stuff cause you are probably heading for disaster. Also consider this, you may be having a problem with reiserfs if you're hd is ok and only the reiserfs gets corrupted. I have switched to jfs because of the problems I had with reiserfs getting corrupted. jfs, as I understand, is the fs used by OS2 which is solid as a rock. Of course, I'm lazy so I look for the easy way out. And being half braindead helps. Richard -- Old age ain't for Sissies!

The Tuesday 2005-03-15 at 17:08 -0000, Kevanf1 wrote:

Don't be tempted to use a brush of any sort to clean the heatsinks. They are like static magnets and will readily discharge against the heatsink.... and of course fry whatever component it is attached to. Either use compressed air or just blow the dust away.

Are you sure of that? Normally the heatsink is isolated, or connected to one of the power rails, or some other pin different from a high impedance input. This is intentional, and makes handling safer. Otherwise, the very same wind generated by the plastic fan blades could build up dangerous static on the heatsink. However, few hobbyist wear a static discharge hand band while working, nor have a controlled humidity. -- Cheers, Carlos Robinson

On Friday 18 March 2005 5:03 am, Kevanf1 wrote:

I will admit that I used to take PC's apart all the time without a static band. But they weren't mine, they belonged to the Uni I worked at. Now, I always put one on when I'm working on my own or friends PC's. I also leave the PC plugged in. Switched off of course :-) ^^^^^^^^^^^^^^^^^^^^^^^ Kevan Farmer

NEVER leave the system plugged in to power. Too many newer systems have Wake-On-LAN or Wake-On-Ring functions with current flowing through the system mainboard and add-in cards. You stand a real chance of damaging the system from AC power this way. AC ground is not the same as grounding yourself to dissipate static electricity. Don't confuse the two, please. Unplug it from the power source and ground yourself and parts to the machine. Its equalizing the static potential that is key. Dissipating static through proper static mats/stations would be even better but that is still equalizing the static charge to all surfaces and components, static bags, etc. Also, compressed air or vacuum cleaners can generate tremendous amounts of static electricity depending on the materials they are made of. Standard plastic vacuum parts are VERY bad for electronic components whether blowing or sucking dust and debris out of a system because they tend to generate static electricity and it IS going to dissipate somewhere. With my luck I don't even THINK about doing this... Sorry if you know this but many people do not so I'd rather warn about hazards than have people let the magic smoke out... Stan

On Sat, 19 Mar 2005 02:08:07 +0100 (CET), Carlos E. R. wrote:

The Friday 2005-03-18 at 16:41 -0600, Stan Glasoe wrote:

...

On Friday 18 March 2005 5:03 am, Kevanf1 wrote:

...

I will admit that I used to take PC's apart all the time without a static band. But they weren't mine, they belonged to the Uni I worked at. Now, I always put one on when I'm working on my own or friends PC's. I also leave the PC plugged in. Switched off of course :-) ^^^^^^^^^^^^^^^^^^^^^^^ Kevan Farmer

NEVER leave the system plugged in to power. Too many newer systems have Wake-On-LAN or Wake-On-Ring functions with current flowing through the system mainboard and add-in cards. You stand a real chance of damaging the system from AC power this way. AC ground is not the same as grounding yourself to dissipate static electricity. Don't confuse the two, please.

I suppose he is referring to leaving the AC cable connected, and the power switch off. I assume that switch is the real, AC, switch, in the power supply box, not the nice button on the front of the PC, that leaves some parts powered.

This way the AC is off, but the ground wire remains connected. Of course, having a separate ground with a 1 Meg resistor would be much better. Safer.

...

Unplug it from the power source and ground yourself and parts to the machine. Its equalizing the static potential that is key. Dissipating static through proper static mats/stations would be even better but that is still equalizing the static charge to all surfaces and components, static bags, etc.

A proper grounding bracelet has a resistor of around 1 meg to limit shock in case of inadvertently touching a live cable. If you are instead well grounded the shock can be lethal.

...

Also, compressed air or vacuum cleaners can generate tremendous amounts of static electricity depending on the materials they are made of. Standard plastic vacuum parts are VERY bad for electronic components whether blowing or sucking dust and debris out of a system because they tend to generate static electricity and it IS going to dissipate somewhere. With my luck I don't even THINK about doing this...

True. But a component connected to its PCB, if it is well designed, should be safe. The idea is that high impedance inputs are connected to low impedance outputs somewhere else, so static can not build up. In theory, that is.

-- Cheers, Carlos Robinson

-- Check the headers for your unsubscription address For additional commands send e-mail to suse-linux-e-help@suse.com Also check the archives at http://lists.suse.com Please read the FAQs: suse-linux-e-faq@suse.com

Sorry, I should have added that yes, the power must be switched off at the wall outlet. NOT just at the front switch. I always set my BIOS to not wake from LAN etc as it is a redundant option for me any way. Though I can see that it may not be for others. Another thought that occurred was different country's power supply options. Here in the UK we have a permanent earth system. We can, if we so wish, pull a power plug partially out of a wall socket disconnecting the live and neutral pins but still leaving the earth pin in place. It's not recommended as it's a bit haphazard but it illustrates a point. Now if I remember rightly, Australia don't have an earth pin? They use two pin round plugs. Or they did 4 years ago when I was working over there. Obviously in this case you couldn't do as I suggested. Here in the UK most places - I have them fitted at home - are now geared up with micro residual current circuit breakers too which trip out at the slightest hint of a short. It probably wouldn't save equipment but it would save your life from an electric shock. -- Take care. Kevan Farmer 34 Hill Street Cheslyn Hay Staffordshire WS6 7HR

On Sun, 20 Mar 2005 13:10:42 -0600, Danny Sauer wrote:

On Sunday 20 March 2005 06:49 am, Kevanf1 wrote:

...

Here in the UK most places - I have them fitted at home - are now geared up with micro residual current circuit breakers too which trip out at the slightest hint of a short. It probably wouldn't save equipment but it would save your life from an electric shock.

Those are called GFCI (ground fault circuit interrupter) outlets in the USA, and are typically only installed near water (in the kitchen and bathroom), as they cost about $7 instead of $0.50 for regular outlets. Many (most?) new houses are built as cheaply as possible over here. :(

--Danny, who generally uses GFCI outlets when feasible

-- Check the headers for your unsubscription address For additional commands send e-mail to suse-linux-e-help@suse.com Also check the archives at http://lists.suse.com Please read the FAQs: suse-linux-e-faq@suse.com

Hmmm, so cost cutting goes on all over :-( We had our house rewired about 7 years ago. Prior to this all the wiring was a right mess with new PVC cabling joined up to ancient black rubber stuff back to PVC. The fuse box was porcelain insulators. We had a complete RCD fuse box protecting every circuit put in. For goodness sake, what price do people put on a life? It reminds me of fellow anglers - I love angling as well as messing with Linux - who go out in a boat with the latest most expensive fishing gear but cannot pay a bit extra for a life jacket. -- Take care. Kevan Farmer 34 Hill Street Cheslyn Hay Staffordshire WS6 7HR

The Sunday 2005-03-20 at 13:10 -0600, Danny Sauer wrote:

Those are called GFCI (ground fault circuit interrupter) outlets in the USA, and are typically only installed near water (in the kitchen and bathroom), as they cost about $7 instead of $0.50 for regular outlets. Many (most?) new houses are built as cheaply as possible over here. :(

They are mandatory in Spain for "new" installations, since perhaps about 20 years ago. The whole house is protected by a "main current limiter" and a "differential switch" in cascade. After that come section limiters, four or six usually (15 to 30A). The differential switch (equivalent to your GFCI) thus protects the whole installation. If one of the two wires carries 20mA more or less than the other (meaning that some is leaking to earth, perhaps through somebody's leg) it switches off everything fast. Of course, our 220V is more dangerous than your 110 (for electrocution, perhaps less for fire), that's why they are mandatory, I guess. So, those things in the movies with the bad guy "dropping" a hair dryer in the bathtub should be useless ;-) On the other hand... do you know that computer's power supplies usually leak a bit to ground? I suppose it is the filter section of switched mode power sources. Sometimes a lab full of PCs can be a nightmare for the GFCI. -- Cheers, Carlos Robinson

Kevanf1 wrote:

On Mon, 21 Mar 2005 14:05:22 +0100 (CET), Carlos E. R. wrote:

Of course, our 220V is more

...

dangerous than your 110 (for electrocution, perhaps less for fire),

Hmmm, forgive me if I'm wrong here - and you guys certainly seem to be far more clued up where electrics are concerned but I have always thought it had nothing to do with the actual voltage but more to do with ampage? So if something was a million volts but only 1 milliamp it would not harm you? Or am I totally off track here?

Take care.

Kevan Cheslyn Hay Staffordshire UK

Sorry that I come in here, but 1 mio volts multiplied by 1 milliamp is still 1 KW, which isn't at all nothing. With only a microamp it might not harm you but I still wouldn't try it ;-) Take care - especially when it comes to high voltage ;-) Martin

The Tuesday 2005-03-22 at 13:33 +0100, Martin Deppe wrote:

Damn' why do I get these (my own) eMails twice (sent)?

I have only seen one copy of each.

I definitely sent it only once.

Could anybody explain that, please?

No, it must be your receiving side of things. You did not send duplicates. You can check your yourself inthe archive (http://lists.suse.com/archive/suse-linux-e/2005-Mar/). -- Cheers, Carlos Robinson

Kevan, On Tuesday 22 March 2005 02:57, Kevanf1 wrote:

On Mon, 21 Mar 2005 14:05:22 +0100 (CET), Carlos E. R. wrote:

Of course, our 220V is more

...

dangerous than your 110 (for electrocution, perhaps less for fire),

Hmmm, forgive me if I'm wrong here - and you guys certainly seem to be far more clued up where electrics are concerned but I have always thought it had nothing to do with the actual voltage but more to do with ampage? So if something was a million volts but only 1 milliamp it would not harm you? Or am I totally off track here?

It's true that current flow is what will harm an animal, but in any given conductive object, applying more voltage drives more current. The relationship between current and voltage is linear, in fact. This is captured by Ohm's law: I = V / R or, in English: Current = Voltage / Resistance.

Take care.

Oh, I will.

Kevan

Randall Schulz

Ken, On Tuesday 22 March 2005 07:08, Ken Schneider wrote:

...

As this thread has now been OT for quite some time can it please be moved to the OT list?

Threads don't move across lists. And off-topicness policing is equally off-topic. Besides, it's not all that off-topic. People who use electrical and electronic devices can benefit from understanding some of these basic concepts.

Ken Schneider

Randall Schulz

Ken, On Tuesday 22 March 2005 08:30, Ken Schneider wrote:

...

...

Besides, it's not all that off-topic. People who use electrical and electronic devices can benefit from understanding some of these basic concepts.

And what does it have to do with the subject as was posted by the OP? Nothing. So in regards to even this thread it is OT.

Threads are conversations. They exhibit topic drift. There's no stopping it and it's pointless to try.

Ken Schneider

Randall Schulz