FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Redhat > Cluster Development

 
 
LinkBack Thread Tools
 
Old 04-14-2008, 04:39 PM
Joel Andres Granados
 
Default extend the --ignoredisk option

Hey list:

What do you guys think about including uuids in the --ignoredisk option? For now I'm thinking of a patch that ignores swap devices based on uuid, but I think it could be extended to whatever uses uuid in anaconda.
This could be a staging point for a new --ignoreswap option or for redefining the --ignoredisk into --ignorestorage (a more general way of defining what needs to be ignored). There are lots of pros and cons for each one. Would like to hear what everyone thinks

Comments appreciated.

--
Joel Andres Granados
Red Hat / Brno, Czech Republic

_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@redhat.com
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
 
Old 04-15-2008, 12:05 AM
Jeremy Katz
 
Default extend the --ignoredisk option

On Mon, 2008-04-14 at 18:39 +0200, Joel Andres Granados wrote:
> What do you guys think about including uuids in the --ignoredisk option? For now
> I'm thinking of a patch that ignores swap devices based on uuid, but I think it
> could be extended to whatever uses uuid in anaconda.
> This could be a staging point for a new --ignoreswap option or for redefining
> the --ignoredisk into --ignorestorage (a more general way of defining what needs
> to be ignored). There are lots of pros and cons for each one. Would like to
> hear what everyone thinks

ignoredisk has been about disks and not partitions, so uuids don't
entirely match up. So having it in the same option seems a little bit
wonky.

What's actually trying to be solved by adding this?

Jeremy

_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@redhat.com
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
 
Old 04-15-2008, 10:19 AM
Joel Andres Granados
 
Default extend the --ignoredisk option

Jeremy Katz wrote:

On Mon, 2008-04-14 at 18:39 +0200, Joel Andres Granados wrote:
What do you guys think about including uuids in the --ignoredisk option? For now
I'm thinking of a patch that ignores swap devices based on uuid, but I think it
could be extended to whatever uses uuid in anaconda.
This could be a staging point for a new --ignoreswap option or for redefining
the --ignoredisk into --ignorestorage (a more general way of defining what needs
to be ignored). There are lots of pros and cons for each one. Would like to
hear what everyone thinks


ignoredisk has been about disks and not partitions, so uuids don't
entirely match up. So having it in the same option seems a little bit
wonky.


Yep, I agree. just thought that having it all in one option would be tidy.




What's actually trying to be solved by adding this?


ignore swap.

What do you think about having --ignoreuuid. This would be really usefull as a way of ignoring whatever has a uuid. It could be a fall back method. when the --ignoredisk cant be used.



--
Joel Andres Granados
Red Hat / Brno, Czech Republic

_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@redhat.com
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
 
Old 04-15-2008, 01:20 PM
Jeremy Katz
 
Default extend the --ignoredisk option

On Tue, 2008-04-15 at 12:19 +0200, Joel Andres Granados wrote:
> Jeremy Katz wrote:
> > What's actually trying to be solved by adding this?
>
> ignore swap.

Ignore it how? Just don't use it during install? Don't add it to the
fstab? What's the use case.

> What do you think about having --ignoreuuid. This would be really usefull as a
> way of ignoring whatever has a uuid. It could be a fall back method. when the
> --ignoredisk cant be used.

And what happens if someone puts in the uuid of a random partition?
What are the semantics of ignoring it -- do we not ever delete the
partition? What happens if they try to use it with ondisk? How does it
interact with zerombr?

ignoredisk is nice and simple -- we remove the disk from the ones we
look and and do anything to. Ignoring just a partition feels far less
obvious to me

Jeremy

_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@redhat.com
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
 
Old 04-15-2008, 02:35 PM
Joel Andres Granados
 
Default extend the --ignoredisk option

Jeremy Katz wrote:

On Tue, 2008-04-15 at 12:19 +0200, Joel Andres Granados wrote:

Jeremy Katz wrote:

What's actually trying to be solved by adding this?

ignore swap.


Ignore it how? Just don't use it during install? Don't add it to the
fstab? What's the use case.


Ignoring on install time. It becomes relative when the installer sees swap partitions that are being used. Might have a disk thats only swap partitions for virtual hosts. If you include that disk in an installation, anaconda will use all the swaps. see 70477.



What do you think about having --ignoreuuid. This would be really usefull as a
way of ignoring whatever has a uuid. It could be a fall back method. when the
--ignoredisk cant be used.


And what happens if someone puts in the uuid of a random partition?
What are the semantics of ignoring it -- do we not ever delete the
partition? What happens if they try to use it with ondisk? How does it
interact with zerombr?

ignoredisk is nice and simple -- we remove the disk from the ones we
look and and do anything to. Ignoring just a partition feels far less
obvious to me



Its not obvious. And to be honest, havent really thought about the interactions with the rest of the ks commands. I would assume that it would have precedence over any of the "modify partitions" options. Thing is, that we would have to check the uuid each time something is being preformed on a storage element. Have to spend some more time pondering that one
We could also code it into the storage classes. So when some partition is ignored, it will have certain behavior that prevents any changes to be done to the underlying structure, but allows data information to be returned to whoever needs it. Then again, it would maybe be asking too much of the storage structure.

I guess I'm just trying to find a way of telling anaconda not to touch certain things.
Darn partitioning code.

I totally agree with you on the complexity, thats the reason for the patch only touches swap related stuff for now. maybe going back to my oritinal --ignoreswap idea is best.




--
Joel Andres Granados
Red Hat / Brno, Czech Republic

_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@redhat.com
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
 
Old 04-16-2008, 02:11 PM
John Summerfield
 
Default extend the --ignoredisk option

Joel Andres Granados wrote:

Jeremy Katz wrote:

On Tue, 2008-04-15 at 12:19 +0200, Joel Andres Granados wrote:

Jeremy Katz wrote:

What's actually trying to be solved by adding this?

ignore swap.


Ignore it how? Just don't use it during install? Don't add it to the
fstab? What's the use case.


Ignoring on install time. It becomes relative when the installer sees
swap partitions that are being used. Might have a disk thats only swap
partitions for virtual hosts. If you include that disk in an
installation, anaconda will use all the swaps. see 70477.


So you're installing to a real disk on a guest system?

I'm not sure I want Anaconda poking around in there.

Any thoughts anyone to installing into a partition (say /dev/sdb2) and
letting the guest use it as a whole disk.


You'd have an mbr in /dev/sdb2 which, one hopes, nobody will try to boot
in the wrong context.




--

Cheers
John

-- spambait
1aaaaaaa@coco.merseine.nu Z1aaaaaaa@coco.merseine.nu
-- Advice
http://webfoot.com/advice/email.top.php
http://www.catb.org/~esr/faqs/smart-questions.html
http://support.microsoft.com/kb/555375

You cannot reply off-list:-)

_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@redhat.com
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
 
Old 04-16-2008, 02:21 PM
Joel Andres Granados
 
Default extend the --ignoredisk option

John Summerfield wrote:

Joel Andres Granados wrote:

Jeremy Katz wrote:

On Tue, 2008-04-15 at 12:19 +0200, Joel Andres Granados wrote:

Jeremy Katz wrote:

What's actually trying to be solved by adding this?

ignore swap.


Ignore it how? Just don't use it during install? Don't add it to the
fstab? What's the use case.


Ignoring on install time. It becomes relative when the installer sees
swap partitions that are being used. Might have a disk thats only
swap partitions for virtual hosts. If you include that disk in an
installation, anaconda will use all the swaps. see 70477.


So you're installing to a real disk on a guest system?


Well its not me, but I'm pretty sure its possible.



I'm not sure I want Anaconda poking around in there.


yep, its not a good idea.



Any thoughts anyone to installing into a partition (say /dev/sdb2) and
letting the guest use it as a whole disk.


The bug specifically talks about using it as a swap disk. Which is not completely inconceivable to me.



You'd have an mbr in /dev/sdb2 which, one hopes, nobody will try to boot
in the wrong context.


If you have a disk, yes. but if you have swap files, this would not be a problem.


--
Joel Andres Granados
Red Hat / Brno, Czech Republic

_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@redhat.com
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
 

Thread Tools




All times are GMT. The time now is 03:14 PM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org