Test inconsistencies between machines on same network

[jolleyutah]

I couldn't see anything in previous posts or the FAQ that was my exact issue.

MacBook Pro - 25/5 Wireless (2.4GHz 802.11n)

Consistent tests of ~25/5 (which is what I would expect, and what other sites show me)

MacPro 1,1 - 10/2 wired @ 1Gb

Consistent tests of ~10/2 (tested in multiple browsers)

MacBookPro - 25/5 Wired (same link/cable as MacPro above)

Consistent tests of ~25/5

Win7 Fusion VM on MacBook Pro - 25/5 Bridged to MBP wireless

Consistent tests of ~25/5

WinXP Fusion VM on MacBook Pro - 25/5 Bridged to MBP wireless

Consistent tests of ~25/5

OSx 10.7 Fusion VM on MacBook Pro - 20/5 Bridged to MBP wireless

OSx 10.7 Fusion VM on MacPro (Same virtual machine as above) - 20/5 Bridged to MacPro en0

My question then is this:

What would cause a single machine to run that markedly slower than all others on the same network, and even than VM's on the same hardware?

I was hoping for consistent "Network" speed testing and not so much weight on the machine doing the testing. My MacPro is my 24/7 machine, and as such I would like to set up automatic testing to gather a couple weeks worth of speed data. But results coming in 60% lower than what I was expecting, and what other machines are testing it concerns me.

Any ideas why a single machine would test so much slower?

MacPro 1,1

Dual Xeon 2.66 dual core

10 GB 667 MHz DDR2 FB-DIMM

OSx 10.7.5

[Pgoodwin1]

I'm a little confused with the configuration of that MacBook Pro. The specs you show above look more like a MacBook Pro 5,3 from 2009. When I look up the specs on The MacBookPro1,1 on MacTracker:

Model Identifier MacBookPro1,1

Introduced January 2006 (Shipped February 2006)

Discontinued May 2006

Order Number MA463LL/A (1.83 GHz), MA464LL/A (2.0 GHz)

Processor Intel Core Duo (T2400, T2500, T2600) ("Yonah")

Processor Speed 1.83, 2.0 or 2.16 GHz

Architecture 32-bit

Number of Cores 2

Cache 2 MB on chip shared L2 cache

System Bus 667 MHz

Benchmarks 1.83 GHz: 2291, 2.0 GHz: 2458, 2.16 GHz: 2660

Storage 80, 100, 120 GB (5400-rpm) or 100 GB (7200-rpm)

Maximum Memory 2.0 GB

Memory Slots 2 - 200-pin PC2-5300 (667MHz) DDR2 SO-DIMM

Ethernet 10/100/1000BASE-T (RJ-45) (support for jumbo frames)

The newest OS that this MacBook Pro can run is 10.6.8. Lion was not supported on the Core Duo processors.

I don't know why it would be that much slower since it does have Gigabit Ethernet, unless that older model isn't connecting as 1000 Base T - maybe it's less forgiving about the cable capacitance. Is the cable a CAT 6 ? if it's a CAT 5, maybe it's reverting back to 100 Base T. But even with 100 Base T it doesn't seem like it would be that much lower....not sure.

The only other speed bottlenecks on that machine would be the 2 Meg RAM limit, and the 80 GB hard drive being very full. But I don't know in your case if either is having an effect. Probably not if only the browser is running.

Disregard the above post. Your model is a Mac Pro not a MacBook Pro. I misunderstood.

The only comment above that still would apply would be about the Gigabit Ethernet and the CAT 5 vs CAT 6 cables because that design was released in 2006.

Pleas disregard my comments above. I was confused with the model. Your issue is with a Mac Pro not a MacBook Pro. Sorry about that.

The only comment I made that would still apply would bee the short discussion on the Ethernet and the CAT 5 and CAT 6 cables.

[jolleyutah]

I understand the confusion.

I thought maybe it was cabling also, so I swapped around several cables, the testing did not follow the cable, but did follow the machine.

Certified Cat6 cable on Slow machine tested fast on fast machine.

Can you elaborate on how the Hard Drive might play into the speed equation?

The fast machine is running SSD, where the slow is running spinning platter (WD black 2tb 64mb cache 7200rpm) slower than SSD, but a pretty fast disk. I also have spinning media on the fast machine (MBP) and VM's running from the slower spinning media test as fast as native OSx on SSD...

[CA3LE]

I understand the confusion.

I thought maybe it was cabling also, so I swapped around several cables, the testing did not follow the cable, but did follow the machine.

Certified Cat6 cable on Slow machine tested fast on fast machine.

Can you elaborate on how the Hard Drive might play into the speed equation?

The fast machine is running SSD, where the slow is running spinning platter (WD black 2tb 64mb cache 7200rpm) slower than SSD, but a pretty fast disk. I also have spinning media on the fast machine (MBP) and VM's running from the slower spinning media test as fast as native OSx on SSD...

That 2TB 7200 RPM drive should be sufficient. You really only see the hard drive speed effecting results when they are much slower. It happens when the hard drive speed is slower than the connection speed because the disk can't keep up.

A good program to test your disk performance in OSX is Blackmagic Disk Speed Test -- free in the app store.

Example, my girlfriends Macbook used to test at ~30Mbps with the original 160GB SATA... I swapped the HDD for a SSD and instantly it was able to max out the 50Mbps connection. Most people wouldn't associate HDD performance with network performance but in some cases it is a factor.

So when you test through VM Fusion on the MacPro (that's getting 10/2 Mbps natively) you're getting your full speed?

[jolleyutah]

Correct. It's almost like there is something specific to that particular install of Lion that is slow.

Perhaps it is time for a rebuild?

Native OSx 10.7.x running on 2tb 7.2k drive = slow

Virtual OSx 10.7.x running on 2TB 7.2k drive = fast

Virtual OSx 10.7.x running on 750gb 7.2k 2.5" Hybrid = fast

[CA3LE]

And you tried using different browsers?