Home Office Network / Switch / VOIP / Cat5e vs Cat6a Help Request

[FreedomPirate01]

Here's my setup:

In one room i have most of the networking equipment. A road runner modem which goes to an airport extreme base station which goes to a new HP J9077A I just got. (Amazing 3x speeds from my qnap 219P to Mac Book Pro over my old dlink switch)
0) To test this i transferred a large folder from Qnap to macbook pro and timed how many MBs transferred in 1 min. Is there an app for osx that i could use instead for more accurate testing in the future?

Then from the HP J9077A i have a cable running to the office where i have 2 more Mac Book Pros hook up via my old dlink router (need to replace). Also a Ricoh Mp 161 is hooked up to the network.

Also from the HP J9077A i have cable running upstairs to the master bedroom to a Pioneer 5020. I'm looking to make this into a media center tv w/ internet use but i haven't had a chance to figure that out yet.

So i have a few questions...
1) Is it better to have the Qnap 219P hooked up directly to the airport extreme base station or the HP J9077A?

I currently have on order a ton of cat 6a patch cables from monoprice to replace all the random cat5 and 5e cables i'm currently using. However, all the inwall cable runs i have in my house currently are cat5e...
2) Should i upgrade some of the more important inwall runs to cat6?

Since i have to upgrade my old dlink router in the office I was thinking of possibily returning the HPJ9077A and buying two HP Switch 1810G - 8 Port. These are the latest and best 8 port switches that hp makes i believe , and should be available in a couple weeks. They are however, smart switches so and i don't' know much about networking so i'm not sure if this is a good idea, though i imagine i could always find someone on craigslist to help me worst case. One feature on the Switch 1810G that looks interesting is EEE 802.3ad Link Aggregation Control Protocol . However, i don't understand how my network would have to be setup to take advantage of this.
3)What kind of setup would i need to take advantage of Link Aggregation?

4) How is performance affected by running 1 cable to my office to a switch (assuming i upgrade my dlink) instead of running 2 or more cables directly to my HP J9077A?

5) Finally, currently i'm using voip phones from timer warner which is hooked up in the basement directly to the coax and the existing POTS phone system, however in the future i might be looking to change to a voip system where the phone themselves would hook directly into my switches. How would such a change affect how I should setup my network? Also how it would affect network performance compared to my currently time warner phone setup?

 

[thiggins]

0) To test this i transferred a large folder from Qnap to macbook pro and timed how many MBs transferred in 1 min. Is there an app for osx that i could use instead for more accurate testing in the future?

I'm not familiar with any.

1) Is it better to have the Qnap 219P hooked up directly to the airport extreme base station or the HP J9077A?

Unless your clients can move traffic at a rate that exceeds 1000 Mbps, then you can put the NAS on either switch.

2) Should i upgrade some of the more important inwall runs to cat6?

Not necessary. CAT5e is fine for gigabit Ethernet.

3)What kind of setup would i need to take advantage of Link Aggregation?

Link aggregation would let you combine ports to increase bandwidth from switch to switch or client to switch. A client would need two Ethernet adapters capable of Link Aggregation. This would come in handy if you had enough clients on either switch to produce more than 1000 Mbps of simultaneous traffic.

4) How is performance affected by running 1 cable to my office to a switch (assuming i upgrade my dlink) instead of running 2 or more cables directly to my HP J9077A?

You can't connect more than one cable from switch to switch unless the switches are capable of link aggregation or spanning Tree (for redundancy). Otherwise you'll get a network storm, with lights on both switches flashing wildly.

5) Finally, currently i'm using voip phones from timer warner which is hooked up in the basement directly to the coax and the existing POTS phone system, however in the future i might be looking to change to a voip system where the phone themselves would hook directly into my switches. How would such a change affect how I should setup my network? Also how it would affect network performance compared to my currently time warner phone setup?

Two VoIP phones won't produce any appreciable traffic on a Gigabit network.

 

[FreedomPirate01]

Testing my network

Thanks for the info Tom. I just read your article on flow control: http://www.smallnetbuilder.com/content/view/30212/55/

...and it made me think whether or not I have this issue.

I transferred a large folder from my Qnap 219p to my macbook pro over my network to test the speed and here are the results i got:

The first time i did this test i transferred a different folder then the next 3 times I did the same folder...

1) 1.34GBs in 1 min
2) 1.80GBs in 1 min
3) 1.60GBs in 1 min
4) 1.63GBs in 1 min

This comes out to about an average of 1.6GBs in 1 min

Please tell me if my understanding is correct....

a Gigabyte network is suppose to run at maximum of 1000Mbps (MegaBits), which is equal to 125MBps (MegaBytes), as 1 MegaByte = 8 MegaBits

So i'm getting 1600MBs in 60 seconds or 26.7 MBps, or in bytes i'm getting 213.333Mbps.

This test I did above is for READ speed on my Qnap, but when i called Qnap they said I should expect about 45MBps READ on the 219P. The NAS charts on this site show it reads at 55MBps.
A) So i'm it looks like there is something wrong with my setup or am I doing the test incorrectly?

B) Assuming I have a flow control issue, what 8port switches would you recommend for my setup?

I just realized in my previous test i included the time it took to "prepare to copy" which i'm assuming now I shouldn't have included....
C) Should i include prep time in my speed tests?

I did some more test without prep time and i got...

1)1.75GB / min
2)1.86GB / min
3)1.72GB / min
4)1.80GB / min
5)1.73GB / min

AVG =1.772GB / min = 29.5 MBps READ on my Qnap 219P

D) ....Still a good bit less than what i should be getting???

 

[thiggins]

Flow control only occurs when a Gigabit source is transferring simultaneously to Gigabit and 100 Mbps clients.

I do not have any benchmarks in the NAS charts for Mac OS clients. My filecopy tests are performed with Vista SP1. The other benchmarks use iozone.

File transfer performance is highly dependent on client OS and on mix of filesizes used.

If you are using jumbo frames, disable them on both client and NAS.

 

[msl]

I transferred a large folder from my Qnap 219p to my macbook pro over my network to test the speed and here are the results i got:
...
So i'm getting 1600MBs in 60 seconds or 26.7 MBps

Is your MacBook actually capable of sustained writing to its internal hard-disk any faster than this? That could easily be your bottleneck.

 

[FreedomPirate01]

Is your MacBook actually capable of sustained writing to its internal hard-disk any faster than this? That could easily be your bottleneck.

Good question... whats the best way to figure that out?

 

[FreedomPirate01]

Upgraded to SS-439 Pro

My 219P wouldn't open up administration page no matter what i tried so i called Newegg even though it was past 30 days and they let me do a full refund. 8)

I ordered a SS-439 Pro and (4) Western Digital Scorpio Black WD3200BEKT 320GB 7200 RPM 16MB Cache 2.5" SATA 3.0Gb/s Internal Notebook Hard Drives. This setup should leave me with about 550GB of free space which should last me for awhile, though it would be nice to have alittle extra room. Hopefully by the time i run out of space they will have bigger drives I can upgrade to.
Total cost including the drives came out to $200 more than my previous 219p setup with (2) 2T drives. Its amazing how little power this setup is supposably going to draw.

I can't wait to see how much faster this setup will be than my previous 219p setup.

 

[Pugs_KDC]

if you have Cat5e inth walls why do you want/need Cat6a patch cables ?? Cat5e is fine for GiGE..... and unless you upgrade the cables in the wall there is no point having Cat6a patch cables.

also no point to cat6a outside of comercial installs that Require 10GiGE copper.... and to that point premade fibre cables are cheaper....


I'm data cabler in aus i for "home office" installs Cat6 is fine

 

[FreedomPirate01]

pugs,
i got the patch cat 6a cables from monoprice.com they were very cheap and not much more than the cat6 or 5e cables so i figured why not be alittle more future proof.

What are these fibre cables you mention?

 

[Jay_S]

Good question... whats the best way to figure that out?

Maybe hdparm from the command line? Does max osx have hdparm (it's a linux util for benching local drives)?
Assuming your drive is "sda",

hdparm -Tt /dev/sda

EDIT
OK, no hdparm that I can see. But:

Test write speed:

time dd if=/dev/zero of=/Volumes/SOMEDISK/testfile bs=1024k count=2048

Where the 2048 should be a number greater than your MB of RAM.

Test read speed:

time dd of=/dev/null if=/Volumes/SOMEDISK/testfile bs=1024k

Source (arstechnica).

 

[Pugs_KDC]

pugs,
i got the patch cat 6a cables from monoprice.com they were very cheap and not much more than the cat6 or 5e cables so i figured why not be alittle more future proof.

What are these fibre cables you mention?

just any premade fibre "patch" cables

 

[FreedomPirate01]

Update on ss 439 pro

So i just did my first test of the qnap ss 439 pro.

everything is the same in the test except i set the jumbo frames to 9k and i'm not sure what the default setting is on the 219p when i did the tests before.

anyways...

I did 2.56GB in 1 min or 42.7MB per second of the same folder to the same computer.

I called Qnap and they I should for SS-439 1 large 5GB file Raid 5 they got 81.4 MB with SAMBA. 55.9 write.

so the only differences between their test and mine should be that i'm using AFP and they transferred one large 5GB file while i transferred music folder filled with 3-4mb mp3 files. So i will do another test with a large file...

 

[FreedomPirate01]

large file test

okay I did a couple more tests and with 1 large file i'm getting 3.9GB per min or 65 MB per second with the SS-439 Pro. Qnap says 81MB / second

also i retested the 219P with a large file and now i'm getting 2.4GB per min or 40MB per second. Qnap says 45MB /second.

Looks like i'm getting closer. The SS 439 seems to be a bit off , but the 219p seems pretty close to what Qnap got.

I wonder what other factors might still be in play...