amuck-landowner

123systems dd on new node.

M

mtwiscool

New Member
heares the dd done over 1 hour on a new(probluly full) node.

about 10x better then old node.

Code: 
root@post2vps:~# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 4.27336 s, 251 MB/s
root@post2vps:~# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 3.30443 s, 325 MB/s
root@post2vps:~# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 3.43913 s, 312 MB/s
root@post2vps:~# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 3.25771 s, 330 MB/s
root@post2vps:~# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 4.01326 s, 268 MB/s
root@post2vps:~# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 3.20705 s, 335 MB/s
root@post2vps:~#
 
GIANT_CRAB

GIANT_CRAB

New Member
lbft said:
Come on, Martin, surely you can squeeze out a proper rant about the good old dd test? :)
Click to expand...
DDs aren't exactly accurate, are they?

OP, try dd without cache or post a serverbear benchmark.

I'm curious about the processing core they are using.
 
Jack

Jack

Active Member
GIANT_CRAB

GIANT_CRAB

New Member
Jack said:
Well what company went bust recently having a shit load of hardware in Dallas with Colocrossing.

http://lowendbox.com/blog/shardhost-13year-20-256mb-ssd-kvm-vps-in-dallas-texas-or-buffalo-ny/

Intel E3 1240v3 and 6 x Intel 520 SSDs with the Raid 10 LSI 9271 8i with Cachevault
Click to expand...
Correct me if I'm wrong...

123system seems to be using Limestone Networks - http://bgp.he.net/ip/64.31.58.10

From what I heard, Limestone Networks isn't very bad, support is pretty good and so are their hardware.

From what I know, Boltersdriveer is hosting with Limestone Networks.
 
Last edited by a moderator:
drmike

drmike

100% Tier-1 Gogent
123system seems to be using Limestone Networks - http://bgp.he.net/ip/64.31.58.10
Click to expand...

Well, from other recent threads on here, 123Systems is moving datacenters.  Limestone is out and CC is where Andrew is going, so it seems.

In the search box up top search all forums for: 123systems 

Look at two most recent threads.

mtwiscool said:
yes ssd caching.

don't know what ssd used.
Click to expand...
If allocated on CC's network, those for now will be Intel 520's.  Every f'n CC company.
 
raindog308

raindog308

vpsBoard Premium Member
Moderator
mtwiscool said:
heares the dd done over 1 hour on a new(probluly full) node.
Click to expand...
Probably new, yes.

Probably full, no.

You'll be able to chart the addition of new VPSes to that server by periodically rerunning your test.  When it falls to sub-megabyte/sec, then the node will be considered 80% full  :p
 
drmike

drmike

100% Tier-1 Gogent
Monkburger said:
Like 'dd' is going to tell you anything...
Click to expand...
With SSD cache in mix, tells you even less useful info.

Someone definitely should revise / come up with way to speed test VPS.  Obviously numbers are only good in aggregate over time as load changes.

Unsure who borrowed tests suitable for bare metal for virtual resources...  
 
Last edited by a moderator:
D

Deleted

Jail
Benchmarking VPS's is utterly silly, interrupts are emulated, containers dont have access to some privileged cpu instructions (and they are emulated), timers are emulated from the host..

The only true way to benchmark a 'VPS' is to run stuff on the host node, when it's under load.. 
 
elohkcalb

elohkcalb

New Member
I for one, still can't figure out why dd result matters when it comes to VPS. Unless I'm doing a lot of writes to the disk all day long, I'm not going to do this dd test any time. However, if I'm doing that much of I/O to the HDD, most likely I will get an abuse warning before anything.
 
Last edited by a moderator:
DomainBop

DomainBop

Dormant VPSB Pathogen
Monkburger said:
Benchmarking VPS's is utterly silly, interrupts are emulated, containers dont have access to some privileged cpu instructions (and they are emulated), timers are emulated from the host..

The only true way to benchmark a 'VPS' is to run stuff on the host node, when it's under load.. 
Click to expand...
All true BUT there are some providers whose marketing strategy revolves around targetting the "VPS Benchmark junkies" segment of the market (wow, look at the dd on that baby, where do I sign up!) so VPS benchmarks may not give an accurate indication of performance but they do have some value as a marketing tool.
 
vRozenSch00n

vRozenSch00n

Active Member
drmike said:
Well, from other recent threads on here, 123Systems is moving datacenters.  Limestone is out and CC is where Andrew is going, so it seems.
Click to expand...
I use their $10/year VPS for NS, just yesterday they informed me about my new IP that changes from 64.31.xxx.xxx by Limestone to 192.227.xxx.xxx by CC.
 
You must log in or register to reply here.
Top
amuck-landowner