zfs discuss - Nov 2008 - Can a zpool cachefile be copied between systems?

Suppose that you have a SAN environment with a lot of LUNs. In the
normal course of events this means that ''zpool import'' is very
slow,
because it has to probe all of the LUNs all of the time.

 In S10U6, the theoretical ''obvious'' way to get around this
for your
SAN filesystems seems to be to use a non-default cachefile (likely one
cachefile per virtual fileserver, although you could go all the way to
one cachefile per pool) and then copy this cachefile from the master
host to all of your other hosts. When you need to rapidly bring up a
virtual fileserver on a non-default host, you can just run
	zpool import -c /where/ever/<host>.cache -a

 However, the S10U6 zpool documentation doesn''t say if zpool cachefiles
can be copied between systems and used like this. Does anyone know if
this is a guaranteed property that is sure to keep working, something
that works right now but there''s no guarantees that it will keep
working
in future versions of Solaris and patches, or something that doesn''t
work reliably in general?

(I have done basic tests with my S10U6 test machine, and it seems to
work ... but I might easily be missing something that makes it not
reliable.)

	- cks

Hello Chris,

Tuesday, November 25, 2008, 11:19:36 PM, you wrote:

CS>  Suppose that you have a SAN environment with a lot of LUNs. In the
CS> normal course of events this means that ''zpool import''
is very slow,
CS> because it has to probe all of the LUNs all of the time.

CS>  In S10U6, the theoretical ''obvious'' way to get around
this for your
CS> SAN filesystems seems to be to use a non-default cachefile (likely one
CS> cachefile per virtual fileserver, although you could go all the way to
CS> one cachefile per pool) and then copy this cachefile from the master
CS> host to all of your other hosts. When you need to rapidly bring up a
CS> virtual fileserver on a non-default host, you can just run
CS>         zpool import -c /where/ever/<host>.cache -a

CS>  However, the S10U6 zpool documentation doesn''t say if zpool
cachefiles
CS> can be copied between systems and used like this. Does anyone know if
CS> this is a guaranteed property that is sure to keep working, something
CS> that works right now but there''s no guarantees that it will keep
working
CS> in future versions of Solaris and patches, or something that
doesn''t
CS> work reliably in general?

CS> (I have done basic tests with my S10U6 test machine, and it seems to
CS> work ... but I might easily be missing something that makes it not
CS> reliable.)

As long as both systems present disks in the same way (same paths) it
will work as expected. Even if they don''t an import should still work
but without the benefit of a cache file.

My understanding is that the the intention behind cachefile property
was to address poor''s man clusters and (hopefully) to enhance
hastoragplus agent
to make use of it automatically so importing a pool would be quicker
in some environments.

The other problem it addresses is that only pools listed in default
cache file will be automatically imported during os boot.



-- 
Best regards,
 Robert Milkowski                            mailto:milek at task.gda.pl
                                       http://milek.blogspot.com

> CS>  Suppose that you have a SAN environment with a lot of LUNs. In the
> CS> normal course of events this means that ''zpool
import'' is very slow,
> CS> because it has to probe all of the LUNs all of the time.
> 
> CS>  In S10U6, the theoretical ''obvious'' way to get
around this for your
> CS> SAN filesystems seems to be to use a non-default cachefile (likely
one
> CS> cachefile per virtual fileserver, although you could go all the way
to
> CS> one cachefile per pool) and then copy this cachefile from the master
> CS> host to all of your other hosts. When you need to rapidly bring up a
> CS> virtual fileserver on a non-default host, you can just run
> CS>         zpool import -c /where/ever/<host>.cache -a
> 
> CS>  However, the S10U6 zpool documentation doesn''t say if
zpool cachefiles
> CS> can be copied between systems and used like this. Does anyone know
if
> CS> this is a guaranteed property that is sure to keep working,
something
> CS> that works right now but there''s no guarantees that it will
keep working
> CS> in future versions of Solaris and patches, or something that
doesn''t
> CS> work reliably in general?
> 
> CS> (I have done basic tests with my S10U6 test machine, and it seems to
> CS> work ... but I might easily be missing something that makes it not
> CS> reliable.)
> 
> As long as both systems present disks in the same way (same paths) it
> will work as expected. Even if they don''t an import should still
work
> but without the benefit of a cache file.
It even works when the nodes have different CTD paths, as the ZFS tries 
to open using devid in case of CTD path failure.

-Venku
> 
> My understanding is that the the intention behind cachefile property
> was to address poor''s man clusters and (hopefully) to enhance
hastoragplus agent
> to make use of it automatically so importing a pool would be quicker
> in some environments.
> 
> The other problem it addresses is that only pools listed in default
> cache file will be automatically imported during os boot.
> 
> 
>