Basil Chupin wrote:
Sid Boyce wrote:
Eberhard Moenkeberg wrote:
Hi,
On Sun, 2 Mar 2008, Juan Erbes wrote:
2008/2/22, Andreas Jaeger
: Sid Boyce
writes: [pruned]
Only 30-40 minutes? Depends of course on the CPU speed and amount of memory. So typically terrabytes is a question of how long is a piece of string.
I love the solidity of ext3. If it does check (wasting your time), you have a far better chance to "survive" against the others.
Viele Grüße Eberhard Mönkeberg (emoenke@gwdg.de, em@kki.org)
Soooooo.... what you are saying here is that ext3 is broke and I should open bugs for 10.3 and 11.0 Alpha2. The 2 instances I related do not indicate solidity. Accidental power loss should only result in replaying the journal. May be in the PC world such things are acceptible, but I could see mayhem if I were depending on a laptop that had to do fsck.ext3 while carrying out maintenance on a mainframe or large SPARC server. I think, really I know for sure that Mr. Customer would seriously become bent out of shape. Luckily using SuSE/reiserfs I have had no such worry or problem over many years, bleeding edge everything and still rock solid.
Reminds me of a laugh British Airways shared with me about Sun's pitch on Concurrent Maintenance which they were touting as a big plus. Scenario put to Sun, box has a hardware hit. Sun says, blacklist the failing hardware, reboot Solaris and they would change the failing bit without powering the system down, so the customer would suffer no downtime. BA says, "but I would suffer a loss of service whether or not you have to power the system down, my mainframes keep on running when they take hardware hits" -- Sun was stuck for an answer. The answer I suppose is down to the amount of grief you can tolerate. Regards Sid.
It seems to me that just because a Linux OS gives a choice of different available filesystems an argument, errr... a heated discussion, must develop.
I never try to get heated, just putting a point of view from another perspective.
If we all were running an M$ system what choice of filesystems would we have?
That is, amongst other reasons, why I removed Windows from my work laptop back in RedHat 6.2 days and never looked back.
In any case, Sid, how many times does one strike a power supply loss? Quite a few times each year, though I'm fortunate to have UPS's installed.
And why would one use a laptop to administer a mainframe or large SPARC server when the power supply in the laptop can go arse-up for no reason?
Ciao.
That's the way it's done, bringing up systems, patch and hardware maintenance, hardware/software installation and configuration. A mobile guy needs a laptop when he's supporting systems everywhere, travelling by car and by plane, a tech support guy needs a heck of a lot of manuals, patches and bulletins on his laptop for quick and easy access when away from the office. Many a time I have NFS mounted my laptop on a Solaris partition and transferred patches across. The laptop scenario is less of a bother where power is concerned, I have had laptops fail, pop the hard drive into another company supplied laptop and soon you have a fully working setup. What was most troublesome was a desktop that suddenly needed a very lengthy fsck.ext3 after an unexplained solid lockup and I could see the same happening with a laptop. So far this has not happened again. Reiserfs seemed (to me) to be more resilient. With a new box here, I am going to give XFS a try later tonight or tomorrow as I am looking for a fs that's is as robust as it can be, JFS looked the part, it had the largest HD and didn't cause the problems I saw with the smaller ext3 partitioned HD. If there are reasons why JFS will be dropped I can accept that. In all the years I've used Linux, ext3 has bitten me more painfully than either reiserfs which I used from SuSE 6.2 onwards and I still have on boxes here. My short excursion with JFS has also been positive. Guys were telling me way back to use ext3 in preference to reiserfs while the kernel mailing list was dealing with a wave of ext3 corruptions. All my reiserfs corruptions then were actually due to failing IDE ports on motherboards (x4). Before anyone suggests Electrostatic Damage, I taught ESD awarenes and I have at my disposal 3 large anti-static mats (about 1.5m x 1m each) properly grounded and I use calibrated Ground Gard wrist strap monitors. 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