On Sun, Nov 07, 2010 at 12:37:25PM -0500, Edward Ned Harvey wrote:
> > From: tech-boun...@lists.lopsa.org [mailto:tech-boun...@lists.lopsa.org]
> > On Behalf Of Derek J. Balling
> >
> > Has anyone done a side-by-side comparison of, say, iSCSI vs. NFS for VM
> > performance?
>
> I did a side-by-si
On Mon, Nov 29, 2010 at 01:55:25PM -0500, Doug Hughes wrote:
> On 11/29/2010 1:43 PM, Paul Graydon wrote:
>> You can actually do 10Gb ethernet over copper, needs to be cat 6A cables
>> though, and it's only good for 100m. By the time you start faffing about
>> with stuff like that I figure the a
we're at an upgrade point on our existing NFS servers and we're starting
to look at our options and I was wondering what experience people have
with Isilon?
We have a fairly small dataset (10TB) but lots of I/O (40-70,000 nfs
ops, with about 10,000 peak 4K writes and 4,000 4K reads).
Thanks!
-j
On Fri, Feb 04, 2011 at 01:27:08PM -0500, Adam Levin wrote:
>
> On Fri, 4 Feb 2011, Jeff Wasilko wrote:
> > we're at an upgrade point on our existing NFS servers and we're starting
> > to look at our options and I was wondering what experience people have
> > wit
On Tue, Mar 08, 2011 at 11:18:15AM -0600, Brodie, Kent wrote:
> I can't speak for other products, but as a relatively new Isilon
> customer, "works for me". Scales well. It's pricey, but you get what
> you pay for. The onefs management interface is decent.
I'd agree Isilon is worth lookin
On Wed, Mar 30, 2011 at 03:53:23PM -0400, Patrick Cable wrote:
> Unfortunately, my existing data is stored on four iSCSI devices chained off
> of a Sun T2000. I seem to be getting 6MB/sec on the rsync, which is painful
> when you're sitting on about 15TB of data.
I've moved 15TB with rsync and I n
On Wed, Mar 30, 2011 at 04:22:24PM -0400, Patrick Cable wrote:
> On Wed, Mar 30, 2011 at 4:00 PM, Jeff Wasilko wrote:
> That could work - i have mulitple $HOME folders (home1, home2, home3...).
> But was your average speed still 6MB/sec? I transferred a couple hundred
> gigs to the
On Wed, Mar 30, 2011 at 04:30:33PM -0400, Singer X.J. Wang wrote:
> My initial theory is that you're CPU bound ATM. The rsync program is a
> single process which cannot use multiple CPUs. I recommend Patrick's idea of
> multiple processes.
IIRC, rsync ends up with multiple processes per job:
>Fro
On Thu, Jun 16, 2011 at 02:21:51PM -0400, John Stoffel wrote:
> The problem with a storage virtualization layer is backups. Or more
> accurately, restores of a complete directory when it's spread across
> multile backends. This was the Acopia model which we tried to make
> work but failed. If yo
On Fri, Mar 23, 2012 at 06:01:35PM +0100, Conrad Wood wrote:
> on storage:
> dd if=/dev/zero of=volume bs=1M : ~ 1,600MByte/s
> cp file1 file2 : ~ 300MByte/s (both files on same volume)
>
> on server
> dd if=/dev/zero of=volume bs=1M : ~ 1,100MByte/s
> cp file1 file2 : ~ 30MByte/s (both files on s
On Tue, Jul 10, 2012 at 09:47:04AM -0700, Andrew Hume wrote:
> any VCS experts?
>
> we want to set up four servers in a single cluster (so that they can
> share
> storage managed by VCFS). we want to configure the 4 servers as two
> active/spare
> pairs. so far, this seems straightforward.
On Tue, Jul 10, 2012 at 09:23:45PM -0700, Gilbert Wilson wrote:
> I'm building out a new data room and wanted to know if anyone has used
> neatpatch.
>
> http://www.neatpatch.com/
I haven't seen their system, but we achieved the same thing using
Ortronics Mighty-Mo racks.
our trick is that we
On Thu, Feb 07, 2013 at 10:46:14AM -0500, Matt Simmons wrote:
> While everyone that I've talked to agrees that both ends need labeled. The
> question is what do you put on them. The schools of thought as far as I am
> aware are:
>
> 1) Every cable end's label says exactly what the other end is con
On Wed, Feb 20, 2013 at 02:16:13PM -0600, Matt Lawrence wrote:
> Believe it or not, this really is a technical question that is sysadmin
> related.
>
> I've just started a new job where I am spending lots of time in a very
> noisy datacenter. Currently I wear earplugs and/or earmuffs. I have tr
On Wed, Apr 03, 2013 at 12:19:04PM -0700, Adrian Luff wrote:
> Depending on the blade manufacturer you're talking about (sounds like HP) the
> likelihood of the passive enclosure backplane causing a failure is near zero.
> We have 4+ years of ~20,000 blades in c-class enclosures without a single
On Wed, Apr 10, 2013 at 10:10:51AM -0500, Matt Lawrence wrote:
> We just received a bunch of 1U servers with cable managment arms (customer
> ordered the arms) and I though I would appeal to the collective wisdom of
> LOPSA. It has been years since I have used cable managment arms and even
> th
On Fri, Jan 23, 2015 at 11:59:09AM -0600, Matt Lawrence wrote:
> Since my mail server is working now, it's time to break my home network
> again, but in a different way.
>
> I'm currently using a Soekris net4801 (running CentOS 4) as a
> firewall/router/DHCP server for my home network and a WRT-
On Fri, Jan 30, 2015 at 08:23:56AM -0800, Tom Perrine wrote:
> With all those drives, you might want to think about using them for
> CrashPlan for your relatives. CrashPlan lets you exchange off site
> back up services with another person. Become "backup buddies" as it
> were.
>
> One scenario -
You could consider an appliance approach...
on the cheap side:
http://www.bswusa.com/Site-Control-Broadcast-Tools-NTP-Server-Sentinel-P8700.aspx?gclid=Cj0KEQjwoYi4BRDF_PHHu6rI7NMBEiQAKZ-JuLzp0zvvbEe8pxNHVIExueJFtjUYir1BjggIv6B-rIEaAs_W8P8HAQ
On te more expensive side:
http://www.endruntechnolog
On Mon, Apr 04, 2016 at 07:42:10PM +, Jeremy Charles wrote:
> Based on earlier feedback from this group, I???ve sent him a potential
> workaround of sticking an NTP server VM on each of four or more physical VM
> hosts. The idea is that if one or two physical hosts get busy, the NTP
> clien
20 matches
Mail list logo
