Re: UFS+J panics on HEAD

From: Steven Hartland <killing_at_multiplay.co.uk>
Date: Thu, 24 May 2012 21:30:17 +0100
On 24/05/2012 00:05, Mark Linimon wrote:
> On Wed, May 23, 2012 at 10:58:48PM +0100, Steven Hartland wrote:
>> > While it might be a shame to see FFS go by the wayside are there any
>> > big reasons why you would rather stick with FFS instead of moving
>> > to ZFS with all the benefits that brings?

>  - ZFS eats bytes for breakfast.  It is completely inappropriate
>    for anything with less than 4GB RAM.
> 
>  - ZFS performs poorly under disk-nearly-full conditions.
>>
>>  - ZFS is not optimal for situations where there are a lot of small,
>>    randomly dispersed IOs around the disk space.  Like in any sort of
>>    RDBMS.

We actually use it in very random high IOP's applications with small
requests, so high standard disk's aren't even an option so SDD's all
the way and we see nice performance.

I can't say we've compared it to say FFS as that simply doesn't provide
the management tools we needed so wasn't even considered, but its far
from shabby in our environment :)

    Regards
    Steve

================================================
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 

In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337
or return the E.mail to postmaster_at_multiplay.co.uk.
Received on Thu May 24 2012 - 18:29:50 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:27 UTC