login
Search
There was a soft updates bug that crept into 4.5-RELEASE, and could cause corruption on shutdown if there was heavy disk activity before the shutdown took place. This hasn't shown up on any lists I read, but it seemed a bit more noteworthy than the latest security hole in some obscure port (no, not PHP).
From the errata page: A bug has been fixed in soft updates that can cause occasional filesystem corruption if the system is shut down immediately after performing heavy filesystem activities, such as installing a new kernel or other software. The system shutdown was unable to flush all buffers on shutdown and would report this fact. The problem can be worked around by running sync(8) a few times before rebooting, or solved by updating to a recent FreeBSD 4.5-STABLE snapshot.
BSD is dying
I think it's "small" issues like this with BSD that keep Linux (and yes, even Microsoft Windows) on top at the end of the day. After all, FreeBSD 4.5-RELEASE is supposed to be exactly that - a release. That means that it's supposed to be used in datacenters, offices and design studios, but at this rate it seems almost unfit for installation on an old "experimentation" box. I'm not trying to put down the BSD people; they've done a great job in the past but it just seems that BSD is a one trick dog - Linux has brought along with it a new standard for open development and collaboration. This bug really should've been found in the testing or unstable branch of FreeBSD and exemplifies perfectly the claims that BSD is dying. IIRC there has never been a single softupdates bug in Linux or indeed in Windows 2000, let alone one that can cause data loss on this scale.
Linux softupdate bug
> IIRC there has never been a single softupdates bug in Linux
That's because there are nothing called ``softupdates" in Linux.
Linux did have problem with sync, mount, ext2, reiserfs, ata,...
Every OS can have bugs.
You shouldn't say the OS is dying only because you see a few bugs.
re: BSD is dying
Don't feed the trolls.
You are absolutly right. Wind
You are absolutly right. Windows 2000 or linux never ever had bug with softupdate... But. Does windows have softupdate. I dont think so.
If you actually use Linux, you sould know that 2.4 kernel which is a release not a dev-kernel wasn't that stable and has a lot of problem with is VM . I don't want to discredit Linux but i'm only saying that you sould aware that every os has bugs even if is a RELEASE or a pair version number.
no it's not correct (was: re: BSD is dying)
i use linux, and i dont like bsd, but what you say is NOT correct!
first, bsd will not die, and i hope that the "free" unix implementations
will keep up, and so everybody can choose wich free os he/she wants.
and second, if you rembember kernel 2.4.15, it had a very similar bug,
linux's filesystem design is different, but the effect of the bug is very similar.
the bug don't hurt my machens because my distro does a tripple sync
before unmount on shutdown, and so nothing happend, i always try the
latest kernels, and that's the price to pay ... bugs will come in
linux, *bsd and other os'es, but we use free operating systems, and
in the worst case we can fix the bugs ourself.
(btw. you compare win2k with bsd? ouch!)
just a opinion of a debian sid (unstable) user ...
BSD still Alive
Recently, Slashdot [goatse.cx] confirmed that WindRiver bucked FreeBSD out on its ass for a carton of Winstons and a 12-pack of Moosehead. This only serves to confirm the fact that FreeBSD is unwanted, doomed to be passed around like an old copy of redhat 7.2.
This is a shame for FreeBSD
The above poster trolled about this, and most of you realized. However, it is too bad that the FreeBSD team didn't get this problem fixed before the release of 4.5-RELEASE. Usually, for a RELEASE version, everythign is supposed to be tested and stable before releasing it, to give end users a stable starting point from which to run a system. I was actually hit by this bug, as my buddy and i were installing FreeBSD for the first time, with the shiny new 4.5 release out, we figured we'd see how it compared to linux. Howver, my buddy and i were pulling our hair out because we'd intall a new kernel, or make a major system change (with a fresh install, we were playing around quite a bit) and upon boot up, we'd get these horrible file system errors, which made no sense. With all the talk of soft-updates versus journalling (which the new linux filesytems use) and comparing of the two, i was interested to see how soft-updates would stack up. Unfortunately, it failed horribly, on even a 'clean' shutdown, much less a 'dirty' shutdown, which is where the new filesystems are supposed to shine. It's really too bad that the FreeBSD team let this error slip through on their latest release system, providing a very bad example of how good their OS really is. And now, as a FreeBSD newbie, i'll have to figure out how to update the system to fix this bug.
Anything titled 'stable' should not have major bugs like this, especially 'release' versions. It's one thing to have bad speed on a particular network card, but major filesystem corruption is awful. For group that releases something like this, it's quite a black eye. It was a black eye for linux, with 2.4.15, and now again for FreeBSD. Let's hope FreeBSD can avoid this in the future, and live up to their reputation of being a 'mature' OS.