[zfs-discuss] ZFS on Linux end-to-end reliability

Johan Hartzenberg johan at webmaster.co.za
Wed Sep 18 09:23:27 EDT 2013


On Wednesday, September 18, 2013 2:31:38 PM UTC+2, Gordan Bobic wrote:
>
> I am not aware of any difference in terms of end-to-end reliability 
> between Linux and Solaris implementations.
>
> There is no such thing as a persistent L2ARC - it needs to get re-heated 
> after a reboot. The data in the ZIL and L2ARC is still checksummed.
>
> I don't know off the top of my head what happens if the block gets 
> corrupted on the ZIL device, perhaps someone else can answer that one.
>

Sorry, I did not mean to focus on the persistent/non persistent aspect of 
the caches, rather to focus on comparing any configuration, eg comparing 
Solaris with Linux when configured with separate ZIL, with separate L2Arc, 
without separate caches, etc.  In short if the reliability is the same 
except when using, for example, separate ZIL, then that is what I would 
like to be aware of.

Really since ZFS on Linux is now considered production ready I want to be 
able to tell the customer that it is "just as reliable as ZFS on Solaris" 
or otherwise I want to recommend Solaris because that is more reliable.
 

To unsubscribe from this group and stop receiving emails from it, send an email to zfs-discuss+unsubscribe at zfsonlinux.org.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.zfsonlinux.org/pipermail/zfs-discuss/attachments/20130918/64198537/attachment.html>


More information about the zfs-discuss mailing list