Good morning,
kernel complains about 30 minutes for some issuses with reiserfs:
Feb 25 06:58:34 kost kernel: [1168633.110949] REISERFS warning (device sda3): jdm-20004 reiserfs_delete_xattrs: Couldn’t delete all xattrs (-39)
Feb 25 06:58:34 kost kernel: [1168633.110949]
Feb 25 06:58:34 kost kernel: [1168633.135718] REISERFS warning (device sda3): jdm-20004 reiserfs_delete_xattrs: Couldn’t delete all xattrs (-39)
Feb 25 06:58:34 kost kernel: [1168633.135718]
I saw this errors like a bug, but in the old reiser 2.6
Server now works, there are no problems. This complains started about 2-3 days after last actualization. ILO says that the server is helathy. There are two SAS disk in the mirros (hardware raid 1). This server is VIP machine, couldn’t reboot it anytime for memtests, checkdisks (sda3 is “/” :-/ ).
Did you saw this anyvhere else ?
Thanks and best regards
J.Karliak.
EDIT: sorry, it’s opensuse 12.3 x64 …(Changed the title to confirm with this)
kernel:
Linux kost 3.7.10-1.28-default #1 SMP Mon Feb 3 14:11:15 UTC 2014 (c9a2c6c) x86_64 x86_64 x86_64 GNU/Linux
reiserfs:
reiserfs-3.6.21-21.1.1.x86_64
I’m still using reiserfs as well. And why should I switch, it never gave me any problems.
I wouldn’t use it on new installations though, I guess…
Thought development pretty much came to a halt after Reiser went to jail for killing his ex.
That happened with reiserfs4.
reiserfs3 was already in the kernel back then (and not developped by Reiser anymore).
And it is still in there and “maintained” by the kernel developers.
But yes, I would also suggest to run “reiserfsck” on it. (you must have the package “reiserfs” installed, not sure if it is installed by default)
I had lots of problems with it I can’t count the times I had to run fsck on it But I admit it was mainly due to power loss. But I have not seen that problem with ext4. Same hardware and occasional power drops.
Hi there,
I never had any problem with reiser (maybe until now ), it survived all forced shutdowns/power loss.
All works fine so far, with the complains. I must wait for time for maintainance or server problems :-/
Anyway, a few years ago I tried btrfs, but thanks to snapshots of the filesystem my / dir became a full. Ok, I can see that now it is a time to learn about btrfs, it is default filesystem in opensuse 13.1 and there is no reiserfs.
Same here.
My root partition uses reiserfs (actually it is one partition over the whole hard disk, including /home), it has not been reformatted since nearly 11 years, and I never needed to run fsck either in all that time.
It always recovered perfectly fine everytime (and I had quite a few situations where it had to in those 11 years… )
But well, YMMV I guess.
There definitely are also people who did have issues with ext4 f.e.
Anyway, a few years ago I tried btrfs, but thanks to snapshots of the filesystem my / dir became a full. Ok, I can see that now it is a time to learn about btrfs, it is default filesystem in opensuse 13.1 and there is no reiserfs.
That’s not true.
Btrfs is not the default filesystem in 13.1 (ext4 still is). And reiserfs is still included, you just cannot select it in the installer anymore.
I am running 13.1 on reiserfs here.
Regarding btrfs: I setup another 13.1 system with btrfs 3 months ago, and so far I quite like it.
But that snapshot feature as it is by default at the moment is a bit problematic IMHO.
While it’s a nice feature in theory, it’s just annoying in practice (especially if you don’t know about it). After 2 months I could not install any updates anymore because the disk was full (although it showed 8 GiB free space). It’s easy enough to delete snapshots in YaST, but how would a Linux newcomer deal with this? (disk full errors although there’s plenty of space…)
So I guess the size of the snapshots should be limited somehow by default, especially if btrfs is really meant to become the new default.
Another (IMHO big) issue at the moment: Grub2 has no write support for btrfs, so if you specify an entry to boot, you will lock yourself out of the boot menu.
This either happens when you choose an entry to boot in KDE’s restart options, run grub2-reboot manually, or even when you hibernate your system. (to prevent data-loss, the default boot entry is set to the hibernated system in that case)
If you choose to boot a parallel installed Windows in KDE’s restart dialog f.e. you even completely lock yourself out of the Linux system and need a LiveCD or similar to recover (because Windows can’t read/write btrfs either of course, so you cannot make the necessary change to get back your boot menu).
See this bug report f.e.: https://bugzilla.novell.com/show_bug.cgi?id=856391
I really think that those 2 issues have to be resolved before btrfs can be made the default filesystem.
Hi,
I’ve fsck’ed /dev/sda3 yesterday evening (after boot to sysrescd) , problem still persists :-/. But now I’ve one of the SAS mirrored disk disconnected from bay. So maybe it is time to begin with changing mirrored disk, one for now.
One question - from the running system I switched to runlevel 1 (single mode). /dev/sda3 is mounted to /var/log but systemd was still running, I couldn’t umount /var/log because of it. It is able to stop this daemon ? Yes, it is system daemon, very very important daemon, but - it is able some?
Thanks and best regards
J.Karliak.
Add “init=/bin/sh” to the boot options. (on grub2 press ‘e’ and append that to the line starting with “linux”, then press ‘F10’ to boot)
You’ll get to a minimal text system then without any init daemon running.
), it survived all forced shutdowns/power loss.
)