Kernel 2.6.8.1 & Reiserfs
Anyone planning to update their kernel to 2. 6. 8 or 2. 6. 8. 1 should be aware that it 'will' trash your Reiserfs volumes. I've had this happen to me, and, I know of at least one other user who had the same happen to him.
Anyone planning to update their kernel to 2.6.8 or 2.6.8.1 should be aware that it 'will' trash your Reiserfs volumes. I've had this happen to me, and, I know of at least one other user who had the same happen to him.
Participate on our website and join the conversation
This topic is archived. New comments cannot be posted and votes cannot be cast.
Responses to this topic
Originally posted by danleff:
Quote:I should have been more clear. Reiserfs 4 is in the YOPER kernel as an option (support). However, if you look at the 2.6.8 changelog, there is a lot of discussion/fixes regarding reiserfs in general.
Anyway, some more discussion about reiserfs 4 is noted on the Yoper forum here.
Interesting. I doubt we're dealing with Reiserfs 4. He may have added support, but, I doubt it. In all probability it's 3.x. Anyway, I've reported it so It will be resolved by the time 2.6.8.x kernel is added to the distro.
Quote:I should have been more clear. Reiserfs 4 is in the YOPER kernel as an option (support). However, if you look at the 2.6.8 changelog, there is a lot of discussion/fixes regarding reiserfs in general.
Anyway, some more discussion about reiserfs 4 is noted on the Yoper forum here.
Interesting. I doubt we're dealing with Reiserfs 4. He may have added support, but, I doubt it. In all probability it's 3.x. Anyway, I've reported it so It will be resolved by the time 2.6.8.x kernel is added to the distro.
I did a little expirementing. I installed yopers latest release on my old PII laptop (which it runs surprisingly well). I used a reiser fs - I don't know what reiser I was using though I know I did use cfdisk for partitioning. I tried a variety of kernels (vanilla, and from yopers own) and the only one that seemed a problem was the reiser + vanilla 2.6.8.1 kernel - yopers latest kernels had no problem - but as of my expirements they had not posted a yoperized 2.6.8.1 kernel to check. The errors I got were
"drive seek complete error sector something er another..."
over and over again on boot - is this the problem that the other people were having??
-So Long and Thanks for all The Fish
"drive seek complete error sector something er another..."
over and over again on boot - is this the problem that the other people were having??
-So Long and Thanks for all The Fish
Originally posted by jimf43:
Quote:Originally posted by trondare:
Quote:Reiserfs is not broken in 2.6.8.1.
I see nothing about this in the kernel mailing lists, forums, kerneltrap, slashdot or anywhere else. I'm using 2.6.8.1-ck1 and reiserfs myself, so how can you say it's broken?
Are you stupid, a troll, or just asking to be flamed?
M'kay. Quick thinking there. I am sure he isn't stupid, mainly because I see no indication of ReiserFS(v3.5/6) problems in the LKML (which there would be).
I myself have seen reiserfs 3.5/6 literally yank out the top level tree upon a reboot. Just becuase I changed kernels. Whereas these 2 kernels have been installed and swapped regular like.
Just because he is challenging your *GLOBAL* announcement the Reiserfs is broken on your machine, therefore it must be broken everywhere, is no reason to give him the STFU TROLL regimen.
Just to give you a clue I had a 3TB array on ReiserFS v3.6 recently "lose" its tree. Was sitting there serving files all fine, when all of a sudden, nearly EVERYTHING was gone.
Brought the spare machine up. Made a 3TB array available to the machine, mkfs.xfs (I wanted to it in production before, but had little chance) as I had tested it and gotten better utilization and performance. Restored. Returned to service in about 30 minutes, with the restore going as the people were using it. Sometimes waiting for files to show up.
The original machine, I ended up trying to fsck it, no good, even some expensive proprietary tools I have couldn't do a thing with it. I ended up having all files under lost+found. Yes, at least I could get to some of them... but the point is: If you aren't prepared for amykind of disaster, you aren't worth keeping around.
Have fun.
Quote:Originally posted by trondare:
Quote:Reiserfs is not broken in 2.6.8.1.
I see nothing about this in the kernel mailing lists, forums, kerneltrap, slashdot or anywhere else. I'm using 2.6.8.1-ck1 and reiserfs myself, so how can you say it's broken?
Are you stupid, a troll, or just asking to be flamed?
M'kay. Quick thinking there. I am sure he isn't stupid, mainly because I see no indication of ReiserFS(v3.5/6) problems in the LKML (which there would be).
I myself have seen reiserfs 3.5/6 literally yank out the top level tree upon a reboot. Just becuase I changed kernels. Whereas these 2 kernels have been installed and swapped regular like.
Just because he is challenging your *GLOBAL* announcement the Reiserfs is broken on your machine, therefore it must be broken everywhere, is no reason to give him the STFU TROLL regimen.
Just to give you a clue I had a 3TB array on ReiserFS v3.6 recently "lose" its tree. Was sitting there serving files all fine, when all of a sudden, nearly EVERYTHING was gone.
Brought the spare machine up. Made a 3TB array available to the machine, mkfs.xfs (I wanted to it in production before, but had little chance) as I had tested it and gotten better utilization and performance. Restored. Returned to service in about 30 minutes, with the restore going as the people were using it. Sometimes waiting for files to show up.
The original machine, I ended up trying to fsck it, no good, even some expensive proprietary tools I have couldn't do a thing with it. I ended up having all files under lost+found. Yes, at least I could get to some of them... but the point is: If you aren't prepared for amykind of disaster, you aren't worth keeping around.
Have fun.
(Synopsis: Problem is occuring on ext3 partitions)
Hey all,
Don't know if this is related or not, but on Friday I upgraded my debian box from a 2.4.25 kernel to a 2.6.8-smp kernel. The former was compiled by me (had to enable the 4GB memory), the latter was installed straight up with the vannilla debian kernel package (binary, smp).
Yesterday morning, the system detected errors in the disk and remounted it read-only, and is now giving read errors all over the place.. however, both partitions are _EXT3_ format, and someone was asking if this happened with such a disk format. The drives are a hardware-mirrored container of two MAXTOR 73GB drives.
I believe someone else mentioned upgrading from a 2.4 kernel before having this problem; perhaps this is relevant? Is there an official location where known bugs/symptoms/causes/fixes/etc will be posted as this problem is catagorized?
Many thanks;
G
While I don't think it's relevant, for the sake of full disclosure (you never know when someone else is going to say 'Hey, I did that too!'), after installing the new kernel (and checking that it worked), I also installed a new raid container of 3 new fujitsu drives, but they haven't been mounted yet anyway.. so like I said, I doubt it's relevant.
Hey all,
Don't know if this is related or not, but on Friday I upgraded my debian box from a 2.4.25 kernel to a 2.6.8-smp kernel. The former was compiled by me (had to enable the 4GB memory), the latter was installed straight up with the vannilla debian kernel package (binary, smp).
Yesterday morning, the system detected errors in the disk and remounted it read-only, and is now giving read errors all over the place.. however, both partitions are _EXT3_ format, and someone was asking if this happened with such a disk format. The drives are a hardware-mirrored container of two MAXTOR 73GB drives.
I believe someone else mentioned upgrading from a 2.4 kernel before having this problem; perhaps this is relevant? Is there an official location where known bugs/symptoms/causes/fixes/etc will be posted as this problem is catagorized?
Many thanks;
G
While I don't think it's relevant, for the sake of full disclosure (you never know when someone else is going to say 'Hey, I did that too!'), after installing the new kernel (and checking that it worked), I also installed a new raid container of 3 new fujitsu drives, but they haven't been mounted yet anyway.. so like I said, I doubt it's relevant.
