Hi Folks - I just was upgraded to 1.5 GbE fibre. (After over 10 years with the Actiontec R1000H my ISP wanted to upgrade the router so bad they gave me a free upgrade from 500/5000 to 1.5/1.5 plus a new free router )
Here is my challenge: The ISP router is about 15-20m (45-45 feet if my math is correct) from my 6100 in my homelab rack. The new router has a 10 GbE copper (RJ45) interface. I have concerns about what I have read on the heat generation from 10 GbE RJ45 SFP+ modules. But, the cost to run optical would involve an RJ45 to SFP+ converter, 2 10 GbE SFP+ optical modules plus, of course, the 15-20m of OM3 which is significantly more expense (tariffs ain’t helpin’ ) than 10 20m Cat6a copper cable plus the RJ45 SFP+ module.
Anyone have any thoughts if the heat from the RJ45 SFP+ module in the 6100 is worth the extra cost?
Actually, after re-reading your post, you are no where near the limit of copper degradation. If I were you I would remap one of the LAN 2.5Gb ports to be your WAN.
This way you can stay on the cheaper side and not have to deal with SFP at all.
Actually, after re-reading your post, you are no where near the limit of copper degradation. If I were you I would remap one of the LAN 2.5Gb ports to be your WAN.
I don’t even think that is necessary. The 6100 has “combo ports” for WAN 1 and WAN 2. I don’t even think any extra configuration is needed. Per the manual its a “use one or the other” type of scenario
Newer / better SFP+ RJ45 10Gb modules use much less power and therefore generate much less heat. Anything less than 3W should be fine from a heat/power perspective in most host devices (Some optical modules can use 2-2.5W, so 3W should be within design margins). An example is the Ubiquiti UACC-CM-RJ45-MG which is rated as 1.9W max and up to 100m, and is going to use less power at shorter distances.
Also you can find used 10g SFP+ modules all over for cheap, the last time I bought some they were $8usd each.
Adding a heatsink to the external portion of the 10gbase-t module has helped me a lot, a simple 14x14x7 adhesive type has lowered temperatures during a benchmark by a noticeable amount. Sorry no before and after real temperatures. I think the bag of these that I bought was around $10usd.
And during a benchmark, the speed doesn’t fall of like it used to do, maybe not be related to the heatsink, but it might be related. Mine is from my n100 based NAS that had a copper port and into my Mikrotik switch which is all SFP+. This module is located in the top row so that the heatsink faces up, not sure if it would be as good in the bottom row.
)
) than 10 20m Cat6a copper cable plus the RJ45 SFP+ module.