Packet Drop Causing 3-5 second delays

[Raist]

Very likely a byproduct of how TCP error detection/correction functions (link is to a simplified explanation, if one is curious).

Basically, an expected segment for a given upload/download is not received in time, and a retransmit is ordered. When your latency is climbing through the roof, it just compounds your lag issue even further due to all the increased communication. It happens more often than people are aware of, but if it happens often enough or with ramped up latency, you may start to see the gap in seconds instead of a light jitter (kind of like a hiccup in a video stream once in a while versus flat out pauses for buffering).

Unfortunately, we may be stuck with this rearing it's ugly head periodically under heavy traffic times. The alternative was probably UDP instead of TCP (like with FFXI), but that creates a whole different problem With UDP, dropped packets would cause blocks of data to just be lost for good and things would simply skip forward on you (like in XI, where people would "blink-walk" around you all the time).

[Coolit]

I've been having packet loss issues for a few days now that eventually cause a 90k error when connected to the EU datacenter. Checked my ISP (Virgin Media UK) status page and they have no issues highlighted for my area, MTR below.

|------------------------------------------------------------------------------------------|

| WinMTR statistics |

| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |

|------------------------------------------------|------|------|------|------|------|------|

| No response from host - 16 | 484 | 409 | 0 | 7 | 23 | 8 |

|sgyl-core-2a-ae8-608.network.virginmedia.net - 25 | 399 | 303 | 5 | 8 | 40 | 10 |

|manc-bb-1d-ae17-0.network.virginmedia.net - 21 | 432 | 344 | 0 | 18 | 68 | 24 |

|leed-bb-1c-ae8-0.network.virginmedia.net - 83 | 182 | 31 | 0 | 16 | 39 | 13 |

|leed-bb-1b-ae10-0.network.virginmedia.net - 74 | 200 | 53 | 0 | 14 | 28 | 16 |

| Lag4.ar9.LON3.gblx.net - 91 | 171 | 17 | 0 | 31 | 36 | 36 |

| ae5.scr3.LON3.gblx.net - 89 | 173 | 20 | 0 | 30 | 54 | 25 |

| po6.ar4.LON3.gblx.net - 87 | 175 | 23 | 0 | 54 | 273 | 25 |

|eidos-vpn.gigabitethernet3-39.ar4.lon3.gblx.net - 91 | 171 | 17 | 0 | 26 | 30 | 29 |

| No response from host - 100 | 157 | 0 | 0 | 0 | 0 | 0 |

| eu.square-enix.com - 5 | 664 | 634 | 25 | 27 | 53 | 26 |

|________________________________________________|______|______|______|______|______|______|

WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud Provider

[teoulas]

Ok the problem has been posted/reported many times already in this forum. It just seems SE doesnt care at all and I think this will be the reason I will stop playing if they dont fix it till end of October.

[Shirai]

|------------------------------------------------------------------------------------------|
|                                      WinMTR statistics                                   |
|                       Host              -   %  | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
|                               TOMODACHI -    0 |  875 |  875 |    0 |    0 |    6 |    0 |
|                           10.15.172.129 -    1 |  871 |  870 |    5 |    8 |   39 |    6 |
|                           212.142.62.89 -    1 |  871 |  870 |    6 |    8 |   35 |    7 |
|      nl-ams05a-rd2-ae-80-2370.aorta.net -    1 |  871 |  870 |    6 |   12 |   62 |    9 |
|                           84.116.138.97 -    1 |  871 |  870 |    7 |   11 |   48 |    8 |
|                          84.116.134.113 -    1 |  871 |  870 |    6 |   10 |   69 |   10 |
|                           213.46.182.18 -    1 |  871 |  870 |    6 |   10 |   66 |    9 |
|            xe-4-2-0.mtl10.ip4.tinet.net -    1 |  860 |  856 |   94 |   99 |  172 |   96 |
|                 ormuco-gw.ip4.tinet.net -    1 |  867 |  865 |   95 |   96 |  119 |   96 |
|                             192.34.76.2 -    1 |  864 |  861 |   95 |   98 |  167 |   95 |
|                          199.91.189.234 -    1 |  864 |  861 |    0 |  101 |  114 |   99 |
|                           199.91.189.38 -    1 |  871 |  870 |   98 |  101 |  113 |  100 |
|________________________________________________|______|______|______|______|______|______|
   WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud Provider

Along the way I have packet loss at pretty much every hop, but per hop this loss is (as far as I can read) within tollerable boundaries, especially compared to a few numbers I see posted here.
Thing is, I have no clue if my readings over all are good or bad as I do experience the "rubber banding" myself from time to time.

Anyone?

[YukiKoneko]

Still an issue...

|------------------------------------------------------------------------------------------|
|                                      WinMTR statistics                                   |
|                       Host              -   %  | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
|          Wireless_Broadband_Router.home -    0 | 1000 | 1000 |    0 |    0 |    1 |    0 |
|    L100.BLTMMD-VFTTP-52.verizon-gni.net -    0 | 1000 | 1000 |    5 |    7 |   35 |    7 |
|  G0-7-2-4.BLTMMD-LCR-21.verizon-gni.net -    0 | 1000 | 1000 |    5 |    9 |   25 |   10 |
|  so-1-1-0-0.RES-BB-RTR1.verizon-gni.net -    1 |  996 |  995 |    7 |   12 |  100 |   10 |
|                0.ae5.XL1.IAD8.ALTER.NET -    7 |  804 |  755 |    0 |   15 |   58 |   15 |
|           0.xe-8-1-0.GW9.IAD8.ALTER.NET -    1 |  996 |  995 |    7 |   11 |   77 |   10 |
|             tinet-gw.customer.alter.net -    2 |  944 |  930 |    8 |   11 |   65 |   12 |
|            xe-4-0-0.mtl10.ip4.tinet.net -    0 | 1000 | 1000 |   25 |   36 |  143 |   93 |
|                 ormuco-gw.ip4.tinet.net -    1 |  996 |  995 |   37 |   41 |   49 |   39 |
|                             192.34.76.2 -    1 |  996 |  995 |   37 |   43 |  105 |   40 |
|                          199.91.189.234 -    1 |  996 |  995 |   42 |   45 |   67 |   47 |
|                           199.91.189.30 -    1 |  996 |  995 |   37 |   41 |   50 |   43 |
|________________________________________________|______|______|______|______|______|______|
   WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud Provider

[Kayko]

Still an issue...
-snip-

Not sure what your trouble is, my loss is 0% -29% - 19% - 66% - 90% - 77% for my 6 hosts..... wasn't even able to silence the first ADS yesterday, and Titans weight kills me before I even see the cast bar.....

[Vapes]

Same problem here, FIOS in Delaware.

[YukiKoneko]

I will say, yesterday, I felt like things ran a bit smoother. Alas, still dropping packets on AENET. The first 200 packets went across fine though...

|------------------------------------------------------------------------------------------|
|                                      WinMTR statistics                                   |
|                       Host              -   %  | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
|          Wireless_Broadband_Router.home -    0 |  500 |  500 |    0 |    0 |    1 |    0 |
|    L100.BLTMMD-VFTTP-52.verizon-gni.net -    0 |  500 |  500 |    5 |    8 |   95 |    7 |
|  G0-7-2-4.BLTMMD-LCR-21.verizon-gni.net -    0 |  500 |  500 |    5 |    8 |   17 |    6 |
|  so-1-1-0-0.RES-BB-RTR1.verizon-gni.net -    0 |  500 |  500 |    7 |   12 |   92 |    9 |
|                0.ae5.XL1.IAD8.ALTER.NET -    2 |  465 |  456 |   13 |   14 |   43 |   13 |
|           0.xe-8-1-0.GW9.IAD8.ALTER.NET -    0 |  500 |  500 |   12 |   16 |   76 |   14 |
|             tinet-gw.customer.alter.net -    1 |  488 |  485 |    7 |   11 |   72 |   17 |
|            xe-4-0-0.mtl10.ip4.tinet.net -    0 |  500 |  500 |   25 |   29 |  128 |   27 |
|                 ormuco-gw.ip4.tinet.net -    0 |  500 |  500 |   40 |   41 |   45 |   42 |
|                             192.34.76.2 -    0 |  500 |  500 |   40 |   43 |  116 |   42 |
|                          199.91.189.234 -    0 |  500 |  500 |   40 |   42 |   76 |   42 |
|                           199.91.189.30 -    1 |  497 |  496 |   40 |   41 |   46 |   41 |
|________________________________________________|______|______|______|______|______|______|
   WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud Provider

[whilke]

Very likely a byproduct of how TCP error detection/correction functions (link is to a simplified explanation, if one is curious).

Basically, an expected segment for a given upload/download is not received in time, and a retransmit is ordered. When your latency is climbing through the roof, it just compounds your lag issue even further due to all the increased communication. It happens more often than people are aware of, but if it happens often enough or with ramped up latency, you may start to see the gap in seconds instead of a light jitter (kind of like a hiccup in a video stream once in a while versus flat out pauses for buffering).

Unfortunately, we may be stuck with this rearing it's ugly head periodically under heavy traffic times. The alternative was probably UDP instead of TCP (like with FFXI), but that creates a whole different problem With UDP, dropped packets would cause blocks of data to just be lost for good and things would simply skip forward on you (like in XI, where people would "blink-walk" around you all the time).

FFXIV uses TCP instead of UDP??? .....

For an endgame that is basically designed for twitch style play, they really didn't think about how to implement a twitch style networking layer.

[toussaint89]

I am taking this that you are all on pc version. Another question is if you guys are using vista-windows 8. If you are a simple fix that fixed this problem for me is Leatrix Latency Fix. Here is the site directly so people don't accidentally download the wrong thing. http://www.leatrix.com/leatrix-latency-fix

For some reason microsoft got this ingenious idea to make your tcp/ip protocol send and receive files at 2-4 at at time. Under heavy traffic times this can cause you to experience delays resulting in you performing an action and nothing happening for seconds at a time. Leatrix fixes this by rewriting the tcp/ip protocols to send and receive 1 file at a time so even under heavy traffic times you don't get that 3-5 second delay.