Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 10 Jul 2012 12:26:06 +0200 (CEST)
From:      Wojciech Puchar <wojtek@wojtek.tensor.gdynia.pl>
To:        Bas Smeelen <b.smeelen@ose.nl>
Cc:        freebsd-questions@freebsd.org
Subject:   Re: FreeBSD 9.0 & LSI MegaRAID SATA problem
Message-ID:  <alpine.BSF.2.00.1207101223170.47120@wojtek.tensor.gdynia.pl>
In-Reply-To: <4FFC013A.3040903@ose.nl>
References:  <CALKKQN=VxL6BzT%2Ba8carbgL_DQK82a91La523E8fo_UuOvNXnA@mail.gmail.com> <op.wg7bf2utg7njmm@michael-think> <alpine.BSF.2.00.1207101051200.46509@wojtek.tensor.gdynia.pl> <4FFBEE6E.6000906@ulb.ac.be> <alpine.BSF.2.00.1207101211030.46945@wojtek.tensor.gdynia.pl> <4FFC013A.3040903@ose.nl>

next in thread | previous in thread | raw e-mail | index | archive | help
>
> That's clear to me.
> These "hardware" raid controllers are not very reliable because they are 
> indeed not real hardware raid controllers, but software based.
> Maybe for desktop usage it's ok/ good enough?

precise what is "desktop" usage is.
i don't see a reason for doing mirroring for home use.

> It is better to use the operating systems raid capability linke gmirror 
> instead.
>
always.

> Of course real hardware raid controllers with cache and battery backed are a 
> different thing and very reliable.

if you have workload where battery backed cache will actually improve 
things (heavy fsync usage) then yes. otherwise no.

i've seen many of them, older, newer, and with same disks i always got at 
least same performance with FreeBSD software solution.

Not talking about RAID5 of which i am not interested at all - there is no 
reason trading performance for available space nowadays with 2-3TB disks.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.BSF.2.00.1207101223170.47120>