ADOBE PREMIERE PRO: Tutorials Forum Articles Creative Cloud Debate

Re: Strategy for Sharing Media Between Mac and Windows Premiere Pro

COW Forums : Adobe Premiere Pro

FAQ   •   VIEW ALL   •   ADD A NEW POST   •   PRINT
Respond to this post   •   Return to posts index   •   Read entire thread


Ed Murphy
Re: Strategy for Sharing Media Between Mac and Windows Premiere Pro
on Dec 6, 2012 at 5:10:46 pm

Hey Petros,

Thanks for your contributions. I wonder if you can speek to my recent experience with link aggregation and managed switches in a Mac environment? I tested and found what I think is strange behavior...

For a few years our setup has consisted of two Netgear GS108T "smart" switches, two Qnap NAS RAIDs for our media library, four MacPros, and two iMacs. The Qnap NASs are each connected to the switches with LACP link aggregation over two ports each.

Now, I've recently found poor network performance on those computers connected via the second switch which I assume is due to the bottleneck of only being uplinked to the first switch via one GbE port. I would like to resolve that, expand network capacity, and setup link aggregation on the four MacPros as you have described, by upgrading our switches. I'm looking at the HP ProCurve you mentioned as well as other Netgear and Cisco options.

However, I wanted to determine what kind of performance boost I could expect before purchasing, so I set up my MacPro with link aggregation and just did a few simple file transfer tests. I believe the results to be strange and I was hoping you might have some thoughts.

Both tests involved copying a 7.14GB file from the Qnap RAID to a local drive and back again. The results were the same in each direction.

Test 1: MacPro connected via single GbE port 5, Qnap connected via aggregated ports 1 and 2. (Current setup.)
Transfer rate in activity monitor peaked and sustained approx 80MB/sec, transfer took 1:41 time.

Test 2: MacPro connected via aggregated ports 3 and 4, Qnap connected via aggregated ports 1 and 2.
Transfer rate in activity monitor peaked around 200MB/sec and sustained approx 160MB/sec, transfer took 1:41 time.

So, while there appeared to be about double the network traffic from one case to the other, the transfer took just as long in both setups! I repeated this several times. How could that be?
Do you have any idea what might be going on?
Should I expect different results with a newer switch, like the HP for example?

Any insights are appreciated.

Ed Murphy
Senior Editor / Technical Director
David Lynch Foundation Television


Posts IndexRead Thread 


Current Message Thread:





© 2019 CreativeCOW.net All Rights Reserved
[TOP]