Control/Dimming Deep questions regarding network switches

smaherIA52

smaherIA52

New Member
Hi all, I've learned through trial and error that Cisco SG-110D-X switches solve problems when they replace Netgear/TP-Link or any and all others. This is all well and good, but I would like to better understand why. I'm working exclusively with ETC/EOS and sACN with only unmanaged switches and static IP's for all gateways. I've only had issues when getting into the teen's and twenty's of universes. It's easy enough for me to say "consumer" vs "enterprise" products but that's a bit too hand wavy for me. Can anyone tell me the real reasons why one fails while the other is solid and worry free? Thanks in advance.
 
You're going to have to be much more detailed about your problems, configuration, equipment, and overall setup to have any useful answers here.
 
danTt said:
You're going to have to be much more detailed about your problems, configuration, equipment, and overall setup to have any useful answers here.
Click to expand...
Thanks for the reply. I've seen fixtures on a rig connected via dmx to eth-dmx gateways that seem to be getting random,noisy data(maybe bitwize offset?) and replacing consumer switches with the CISCO's will solve the problem. I've seen this with many types of fixtures as well as laggy communication between multi-console setups. I can solve the problem and get the work done with "better" switches, I'd just like to have a better understanding of what the actual problem is(and yes all switches were gigabit both good and bad). Thanks again.

also:
RDM disabled in the desk and gatewys
All protos off save for sACN
Single network with /16 mask
 
Last edited:
It could be because of the backplane rating for maximum data handling. A 48 port 10/100/1000 switch can not usually handle 48000bps. Just a guess. Also, I believe unmanaged switches in a sACN system are handling more data throughput than managed switches using IGMP.
John
 
I've been using Art-net since its inception (1998), and sACN since it's availability, on shows exceeding 200 universes. I have never had issues with netgear/TP-link/linksys etc...
(The exception is that some "smart switches" , regardless of brand see art-net and sACN as a ddos attack and try to block it.)

Art-net 1, which was designed for 10Mbs connections can easily handle 40 universes, so if you are getting issues on "teen's and twenty's of universes" over a 1000Mbs network, something other than switch manufacturer (or the backplane) is the issue, especially with the less bandwidth intensive sACN.

"bitwise offset" or noisy data should not be possible with sACN or Art-Net,

Unless you are setting up advanced topologies, the only advantage enterprise switches should give you is long term hardware reliability and service agreements.
As an aside, entertainment specific switches give the most appropriate feature set for our business. Pathway, Luminex, ELC, Netgear ProAV, TMB and others all will beat Cisco/Aruba/Juniper/Arista for our market for price to feature ratios.

Good luck.
 
John Palmer said:
It could be because of the backplane rating for maximum data handling. A 48 port 10/100/1000 switch can not usually handle 48000bps. Just a guess. Also, I believe unmanaged switches in a sACN system are handling more data throughput than managed switches using IGMP.
John
Click to expand...
"teen's and twenty's of universes" are going to use less than 5Mbs on a 1000Mbs network... backplane will not be an issue.
Additionally, such a small universe count shouldn't even saturate old nodes where IGMP snooping would be of benefit.
A quick glance at the listed cisco switches data sheet does indicate that there is a loop protection enabled by default, (maybe RSTP or somethin Cisco proprietary, ) that other switches may not have, or may not be enabled by default... Only thing that would explain a difference in performance.
 
I'm always a little suspicious of unmanaged switches that have "smart" features like Green/Energy-Efficient Ethernet. There are a lot of factors that could play into whether or not that protocol can cause a problem, but it's at least theoretically possible for it to cause some unexpected latency and that could account for the weird behavior. It's not that actual values are wrong, but that they aren't arriving everywhere at the right time. I've never spent much time investigating what it takes to make that happen, as I've always just bought switches that have the ability to turn it off, and then done so.

I do have to disagree on the concept that entertainment-specific switches are necessarily a better value, though. You can get a more capable switch for less money from Cisco than you can from Pathway (for example). However, that Cisco switch has a steeper learning curve in configuring it properly. The advantage of the Pathway (and others) switches is in the configuration process--not its inherent quality or capability. If you're looking at something you're going to reconfigure frequently, that certainly has value. If it's going to be shoved in a rack and ignored for 5-10+ years until it breaks, then there's less of an obvious advantage.
 
RonaldBeal said:
I've been using Art-net since its inception (1998), and sACN since it's availability, on shows exceeding 200 universes. I have never had issues with netgear/TP-link/linksys etc...
(The exception is that some "smart switches" , regardless of brand see art-net and sACN as a ddos attack and try to block it.)

Art-net 1, which was designed for 10Mbs connections can easily handle 40 universes, so if you are getting issues on "teen's and twenty's of universes" over a 1000Mbs network, something other than switch manufacturer (or the backplane) is the issue, especially with the less bandwidth intensive sACN.

"bitwise offset" or noisy data should not be possible with sACN or Art-Net,

Unless you are setting up advanced topologies, the only advantage enterprise switches should give you is long term hardware reliability and service agreements.
As an aside, entertainment specific switches give the most appropriate feature set for our business. Pathway, Luminex, ELC, Netgear ProAV, TMB and others all will beat Cisco/Aruba/Juniper/Arista for our market for price to feature ratios.

Good luck.
Click to expand...
Thanks Ronald, Everything you say makes sense to me and has been my understanding for years, and yet empirically I get other results. I guess I need to start looking closer at the real traffic with wireshark or similar and see what else I can learn.
Thanks for the reply.
 
Malabaristo said:
I'm always a little suspicious of unmanaged switches that have "smart" features like Green/Energy-Efficient Ethernet. There are a lot of factors that could play into whether or not that protocol can cause a problem, but it's at least theoretically possible for it to cause some unexpected latency and that could account for the weird behavior. It's not that actual values are wrong, but that they aren't arriving everywhere at the right time. I've never spent much time investigating what it takes to make that happen, as I've always just bought switches that have the ability to turn it off, and then done so.

I do have to disagree on the concept that entertainment-specific switches are necessarily a better value, though. You can get a more capable switch for less money from Cisco than you can from Pathway (for example). However, that Cisco switch has a steeper learning curve in configuring it properly. The advantage of the Pathway (and others) switches is in the configuration process--not its inherent quality or capability. If you're looking at something you're going to reconfigure frequently, that certainly has value. If it's going to be shoved in a rack and ignored for 5-10+ years until it breaks, then there's less of an obvious advantage.
Click to expand...
I have definitely run into "smart" unmanaged switches or those with "convienient web interface" that were an immediate problem. I've also come to like any dmx-gateway that includes a second eth port and eliminates the need for a separate switch in many cases. Many Thanks
 
smaherIA52 said:
Thanks Ronald, Everything you say makes sense to me and has been my understanding for years, and yet empirically I get other results. I guess I need to start looking closer at the real traffic with wireshark or similar and see what else I can learn.
Click to expand...
If you know wireshark exists you're three steps ahead of most civilians, so... :)
 
If it's a "basic" layer two switch, then likely it helps that it's not overthinking the data. It sees the destination and sends it, no special rules or bloatware interference.

Layer three switches, same thing, with the bonus that it can figure out more advanced routing.

(I'd be more specific but my Cisco training is a bit old now)
 
You must log in or register to reply here.

Similar threads

M
DMX network for new led lights and movers
Replies
11
Views
2K
StradivariusBone
StradivariusBone
joshhealing
Help Needed: No idea how school's system works.
Replies
16
Views
4K
NickVon
NickVon
K
QLab OSC & EOS Ti missing cues
Replies
7
Views
2K
almorton
A
gafftaper
Design The New Pathway eLink and my lighting system upgrade
Replies
1
Views
1K
tyler.martin
T
Kyla Werrett
A.C. Lighting Inc. Delivers a Stellar Lighting Rig to Discovery Church
Replies
1
Views
1K
RonHebbard
R

Users who are viewing this thread

Total: 1 (members: 0, guests: 1)
Back