Mirror of Raidz for data reliability

[Marcelo Araujo]

2012/5/19 Daniel Kalchev <daniel at digsys.bg>

>
>
> On 18.05.12 19:55, Trent Nelson wrote:
>
>> So, my thinking isŠ because both machines can see all disks, the master
>> could import the zpool as normal, and the slave could import it read-only.
>> (Or not import it at all...)
>>
>
> The proper way of doing it is "not import it at all". ZFS is not an shared
> filesystem.
>
> If you have the second host mount the zpool even if read-only, you only
> guarantee that data on the pool will not be corrupted, but you cannot avoid
> the second "read-only" host panic or otherwise crash if it tries to access
> data which is no longer where it thinks it is, because the second host
> doesn't have access to the primary host's in-memory metadata about ZFS.
> Since ZFS is copy on write filesystem, chances are you will be accessing
> data that is no longer valid. Refreshing the internal ZFS state between two
> or more hosts is non-trivial (if it was, Sun would have done this, as it
> suits their usage) and in any case performance will suffer at least as much
> as an true networked filesystem does, compared to "native" ZFS.
>
>

Yeap, you are right! In my environment I'm not doing ACTIVE-ACTIVE servers,
so, one of those controller always will be in stand-by mode. In case
someone need to have all data available in both machines at the same time,
the best choice right now is use HAST.

But my solution is to reach another necessity.

Best Regards,
-- 
Marcelo Araujo
araujo at FreeBSD.org