Date: Tue, 4 Apr 1995 14:59:55 -0700 (PDT) From: "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com> To: gibbs@estienne.CS.Berkeley.EDU (Justin T. Gibbs) Cc: hackers@FreeBSD.org Subject: Re: large filesystems/multiple disks [RAID] Message-ID: <199504042159.OAA08565@gndrsh.aac.dev.com> In-Reply-To: <199504042140.OAA12675@estienne.cs.berkeley.edu> from "Justin T. Gibbs" at Apr 4, 95 02:40:04 pm
next in thread | previous in thread | raw e-mail | index | archive | help
[CC: trimmed] > > >> > >> > RAID does have the negative effect of of having to write 20% more data, > >> > thus cutting effective bandwidth by 20%. It is actually worse than > >> > this in that all writes must write to at least 2 drives no matter how > >> > small they are. The removes some of the benifits of stripping. > >> > >> And that is why some RAID systems use (battery backed up please ;-) RAM > >> caches. This works quite nicely. > > > >And you find these caches will fill up and some point in a sustained > >write test and you end up right back at the 20% performance loss I > >was talking about. > > Is this still true with hardware parity calculation? It's not the time to caclulate the parity that hurts you, it is the fact that you have to write that 20% extra data some place, and that some place had better not be the drive you wrote the real data to! So in effect all write to a RAID system have to write 2 drives at the same time, meaning for a RAID 5 device your speed improvement is on the order of (N/2) or 5/2 while on a pure stripe system you will get an order of N speed improvement. > >Pure stripping of drives always outperforms RAID, you always pay some > >price for reliability, and it is usually performance or $$$. > > > -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Custom computers for FreeBSD
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199504042159.OAA08565>