save/restore problems

save/restore problems

[Hari Kodungallur]

Hi All,

I am running into a bunch of problems when trying to do save/restore.
(I am NOT running the latest xen. I am running 2.0)

I am trying out the FC-1, and RH-ES-9 images. I can create domains for
all the three. But when I save some configurations of these images
(for example, install some rpms on a clean image and save the image)
and then restore them, I am running into the following issues:

(1) FC-1: The most success I have is with this. I can do save and
restore. But as soon as I execute a command remotely that modifies the
file system (e.g., "ssh -n hostname "do-some-fs-updates.sh"), it
starts complaining that the filesystem is read only. The error on the
console looks something like:

xen_blk:  Unexpected blkif status disconnected in state connected
blkfront: recovered 0 descriptors
nfs warning: mount version older than kernel
nfs warning: mount version older than kernel
EXT3-fs error (device sda1): ext3_free_blocks: bit already cleared for
block 293633
Aborting journal on device sda1.
ext3_abort called.
EXT3-fs error (device sda1): ext3_journal_start: Detected aborted journal
Remounting filesystem read-only
EXT3-fs error (device sda1) in start_transaction: Journal has aborted
ext3_reserve_inode_write: aborting transaction: Journal has aborted in
__ext3_journal_get_write_access<2>EXT3-fs error (device sda1) in
ext3_reserve_inode_write: Journal has aborted
ext3_reserve_inode_write: aborting transaction: Journal has aborted in
__ext3_journal_get_write_access<2>EXT3-fs error (device sda1) in
ext3_reserve_inode_write: Journal has aborted
EXT3-fs error (device sda1) in ext3_orphan_del: Journal has aborted
EXT3-fs error (device sda1) in ext3_truncate: Journal has aborted
EXT3-fs error (device sda1) in start_transaction: Journal has aborted
EXT3-fs error (device sda1) in start_transaction: Journal has aborted


Doing this a couple of times (shutdown and then restore again)
corrupts the fs and then i need to run fsck at boot time to get back
to normal status.


(2) RH-ES-9: The save command ("xm save RHES9 myRHES9") just hangs.
Does not do anything.


My question is whether this is something that anyone have seen before
and/or whether anyone could point me to why this happens here and/or
whether installing 2.0.5 would solve this for me.

Thanks
-Hari


-------------------------------------------------------
This SF.net email is sponsored by Demarc:
A global provider of Threat Management Solutions.
Download our HomeAdmin security software for free today!
http://www.demarc.com/Info/Sentarus/hamr30

Fwd: save/restore problems

[Hari Kodungallur]

Thanks, B.

Mark Williamson replied earlier today that the second problem could
have been because of known issues that have since been fixed and the
first issue was because of something else that I might need to help
investigate why.
If that is the case, could you or Mark guide me as to how to go about
debugging the problem. If you can give me some pointers, I will take
it from there.

Thanks very much for your responses.

(I had earlier replied to Mark's message, but to the old list --
xen-devl@lists.sourceforge.net)

Thanks again
-Hari


---------- Forwarded message ----------

This is a known issue fixed in later releases.  BTW, the list has moved,
you might want to re-register ....

B.


On Wed, 2005-03-30 at 00:51, Hari Kodungallur wrote:
> Hi All,
>
> I am running into a bunch of problems when trying to do save/restore.
> (I am NOT running the latest xen. I am running 2.0)
>
> I am trying out the FC-1, and RH-ES-9 images. I can create domains for
> all the three. But when I save some configurations of these images
> (for example, install some rpms on a clean image and save the image)
> and then restore them, I am running into the following issues:
>
> (1) FC-1: The most success I have is with this. I can do save and
> restore. But as soon as I execute a command remotely that modifies the
> file system (e.g., "ssh -n hostname "do-some-fs-updates.sh"), it
> starts complaining that the filesystem is read only. The error on the
> console looks something like:
>
> xen_blk:  Unexpected blkif status disconnected in state connected
> blkfront: recovered 0 descriptors
> nfs warning: mount version older than kernel
> nfs warning: mount version older than kernel
> EXT3-fs error (device sda1): ext3_free_blocks: bit already cleared for
> block 293633
> Aborting journal on device sda1.
> ext3_abort called.
> EXT3-fs error (device sda1): ext3_journal_start: Detected aborted journal
> Remounting filesystem read-only
> EXT3-fs error (device sda1) in start_transaction: Journal has aborted
> ext3_reserve_inode_write: aborting transaction: Journal has aborted in
> __ext3_journal_get_write_access<2>EXT3-fs error (device sda1) in
> ext3_reserve_inode_write: Journal has aborted
> ext3_reserve_inode_write: aborting transaction: Journal has aborted in
> __ext3_journal_get_write_access<2>EXT3-fs error (device sda1) in
> ext3_reserve_inode_write: Journal has aborted
> EXT3-fs error (device sda1) in ext3_orphan_del: Journal has aborted
> EXT3-fs error (device sda1) in ext3_truncate: Journal has aborted
> EXT3-fs error (device sda1) in start_transaction: Journal has aborted
> EXT3-fs error (device sda1) in start_transaction: Journal has aborted
>
>
> Doing this a couple of times (shutdown and then restore again)
> corrupts the fs and then i need to run fsck at boot time to get back
> to normal status.
>
>
> (2) RH-ES-9: The save command ("xm save RHES9 myRHES9") just hangs.
> Does not do anything.
>
>
> My question is whether this is something that anyone have seen before
> and/or whether anyone could point me to why this happens here and/or
> whether installing 2.0.5 would solve this for me.
>
> Thanks
> -Hari

RE: save/restore problems

[Ian Pratt]

'xm save' only saves the memory image of the domain. You should also
save a snapshot of the domain's file system using LVM (lvcreate -s) if
you intend to modify it. 

Restoring a domain if the file system has changed behind its back is
sure to lead to disaster (unless you're using NFS).

Ian

> Mark Williamson replied earlier today that the second problem 
> could have been because of known issues that have since been 
> fixed and the first issue was because of something else that 
> I might need to help investigate why.
> If that is the case, could you or Mark guide me as to how to 
> go about debugging the problem. If you can give me some 
> pointers, I will take it from there.
> 
> Thanks very much for your responses.
> 
> (I had earlier replied to Mark's message, but to the old list --
> xen-devl@lists.sourceforge.net)
> 
> Thanks again
> -Hari
> 
> 
> ---------- Forwarded message ----------
> 
> This is a known issue fixed in later releases.  BTW, the list 
> has moved, you might want to re-register ....
> 
> B.
> 
> 
> On Wed, 2005-03-30 at 00:51, Hari Kodungallur wrote:
> > Hi All,
> >
> > I am running into a bunch of problems when trying to do 
> save/restore.
> > (I am NOT running the latest xen. I am running 2.0)
> >
> > I am trying out the FC-1, and RH-ES-9 images. I can create 
> domains for 
> > all the three. But when I save some configurations of these images 
> > (for example, install some rpms on a clean image and save 
> the image) 
> > and then restore them, I am running into the following issues:
> >
> > (1) FC-1: The most success I have is with this. I can do save and 
> > restore. But as soon as I execute a command remotely that 
> modifies the 
> > file system (e.g., "ssh -n hostname "do-some-fs-updates.sh"), it 
> > starts complaining that the filesystem is read only. The 
> error on the 
> > console looks something like:
> >
> > xen_blk:  Unexpected blkif status disconnected in state connected
> > blkfront: recovered 0 descriptors
> > nfs warning: mount version older than kernel nfs warning: mount 
> > version older than kernel EXT3-fs error (device sda1): 
> > ext3_free_blocks: bit already cleared for block 293633 Aborting 
> > journal on device sda1.
> > ext3_abort called.
> > EXT3-fs error (device sda1): ext3_journal_start: Detected aborted 
> > journal Remounting filesystem read-only EXT3-fs error 
> (device sda1) in 
> > start_transaction: Journal has aborted
> > ext3_reserve_inode_write: aborting transaction: Journal has 
> aborted in 
> > __ext3_journal_get_write_access<2>EXT3-fs error (device sda1) in
> > ext3_reserve_inode_write: Journal has aborted
> > ext3_reserve_inode_write: aborting transaction: Journal has 
> aborted in 
> > __ext3_journal_get_write_access<2>EXT3-fs error (device sda1) in
> > ext3_reserve_inode_write: Journal has aborted EXT3-fs error (device 
> > sda1) in ext3_orphan_del: Journal has aborted EXT3-fs error (device 
> > sda1) in ext3_truncate: Journal has aborted EXT3-fs error (device 
> > sda1) in start_transaction: Journal has aborted EXT3-fs 
> error (device 
> > sda1) in start_transaction: Journal has aborted
> >
> >
> > Doing this a couple of times (shutdown and then restore again) 
> > corrupts the fs and then i need to run fsck at boot time to 
> get back 
> > to normal status.
> >
> >
> > (2) RH-ES-9: The save command ("xm save RHES9 myRHES9") just hangs.
> > Does not do anything.
> >
> >
> > My question is whether this is something that anyone have 
> seen before 
> > and/or whether anyone could point me to why this happens 
> here and/or 
> > whether installing 2.0.5 would solve this for me.
> >
> > Thanks
> > -Hari
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
>