BackupPC-users

Re: [BackupPC-users] serial or parallel backups

2011-06-27 16:42:51
Subject: Re: [BackupPC-users] serial or parallel backups
From: Dan Pritts <danno AT internet2 DOT edu>
To: "General list for user discussion, questions and support" <backuppc-users AT lists.sourceforge DOT net>
Date: Mon, 27 Jun 2011 16:40:56 -0400
> 
> What is your i/o subsystem? I have a striped array (raid 0) over two
> raid 6 arrays with 7 drives in each array, so effectively I have 10
> spindles. With this I can handle more i/o load than if I only had one
> drive.

I'm sure that John knows this, but for the benefit of the OP i'll note that 
it's a lot more complicated than that, of course.  For small writes I might say 
that you effectively have 2 spindles, and since each small write needs two disk 
operations I might even say you effectively have only 1 spindle.

OTOH for small reads you have 14 spindles.  

I don't mean to start an argument, and the numbers above are basically pulled 
out of my butt to make my point.  I just want to note for the OP that it gets 
complicated and you should do more research.  

> I would expect the throughput of multiple servers backing up in
> parallel to be less the the total i/o bandwidth of the disk since each
> write will end up moving the heads of the disks to different locations
> dropping the effective bandwidth of the disk.

Multiple I/O streams can probably make better use of your I/O bandwidth than 
can a single serial one, unless you are running on low-end hardware like a USB 
drive.

The OS and/or RAID controller and/or disks will reorder requests that come in 
in parallel so they can be more efficiently performed.  

Good enough hardware would be a SATA disk with NCQ enabled.  NCQ is a function
of the disk AND the controller AND the OS, don't assume that all SATA does it.

danno
--
Dan Pritts, Sr. Systems Engineer
Internet2
office: +1-734-352-4953 | mobile: +1-734-834-7224


------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security 
threats, fraudulent activity, and more. Splunk takes this data and makes 
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
_______________________________________________
BackupPC-users mailing list
BackupPC-users AT lists.sourceforge DOT net
List:    https://lists.sourceforge.net/lists/listinfo/backuppc-users
Wiki:    http://backuppc.wiki.sourceforge.net
Project: http://backuppc.sourceforge.net/