OptimumOnline → Re: Finally got bufferbloat to A

Those look pretty good to me. All 4 of the traces show about 12ms until they leave the OOL network. There does not seem to be any real packet loss, the 100% hops just are configure to ignore ICMP. Your line quality per the dslreports speed test was an A.

Most of the increase in round trip time seems to be on the cogento network, so it is likely specific to that route to dslreports. It will not change the overall performance of your connection and even that increase is not a real problem.

So unless you have a specific problem elsewhere, you are good to go.

Thanks you Sir, I'm still new to the service just keeping an eye on things. I appreciate your time, thanks again.I not disagreeing with you, the point is with house wiring, patch panel, Ethernet switch, station cabling and router removed there was an slight improvement with buffer bloat (not speed) when connected directly to the ONT. The flipside is when connecting to the OOL modem either by a direct connection or going through my home network I still get an F buffer bloat rating, it make no difference. So is the buffer bloat just a “feel good” measurement as Irish Shark may have eluded to? I don’t know, I'm just looking to optimize what I have currently

IMHO, *IF* you are pushing your connection to the limit (Gaming via Steam, Multiple, High-bandwidth Downloads, Streaming HD Video, etc.) then Bufferbloat *IS* a factor to be considered...

(Edit): Removed Non-Applicable statement, since you posted your results with QoS enabled.

Those two devices have buffers, so it doesn't really have anything to do with the "wiring" as much as the devices in the path. The router would likely contribute more to Bufferbloat than the switch, but I suppose both could affect the rating.I don't know how justin has set up the scale for grading, so it's possible the "clean" FiOS connection is a borderline "C" and adding the additional equipment to the mix just slightly pushes into over to the next grade. While a "clean" OOL connection may already be firmly planted in the "F" range, making the in-home equipment a non-factor.

As for this discussion on if it matters, granted my knowledge is limited, but I don't think any of the applications brought up so far would really benefit from a Bufferbloat score of "A" on this test unless you were already experiencing problems with those applications or you are consistently maxing out your connection.

Gaming is actually a fairly low-bandwidth application, as are video conferencing (by today's standards), VoIP, etc. These applications on their own may not induce bufferbloat while the way the DSLR test is performed will. Latency is important and a good bufferbloat score will ensure good latency at the expense of total throughput, but the way the test measures bufferbloat is not the only way this problem manifests itself.

In other words, if you are maxing out your connection on other things while trying to play games or video conference, then yes, getting an "A" will help you. But if you are not experiencing the affects of Bufferbloat without maxing out your connection and you're not downloading the latest torrent while on a conference call, then "fixing" it won't really change anything.

I think the service is currently running at optimum performance no pun indented. Due to seeing some faulty modem stats a few weeks ago I may be now paying a little more attention to the service than required. So that’s another reason why I asked TheWiseGuy to view my Pingplotter results, I don’t fully understand them and it’s always good to get another view. As far as the buffer bloat I improved ours as detailed above.. My view is if the DSLreports mods thought buffer bloat was important enough to include as a benchmark why not try and correct it especially since we do stream...