dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
1696
share rss forum feed


Baud1200

join:2003-02-10
Reviews:
·Shaw

Shaw Peering to Tata communications = insane packet loss..

|---------------------------------------------------------------------------------- --------|
 
|                                      WinMTR statistics                                   |
 
|                       Host              -%Loss | Sent | Recv | Best | Avrg | Wrst | Last |
 
|------------------------------------------------|------|------|------|------|------|------|
 
|                DarkHorse.PrisonProvince -    0 |   45 |   45 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                           64.59.xxx.xxx -    0 |   45 |   45 |    8 |   14 |   26 |   15 |
 
|                            66.163.78.xx -    0 |   45 |   45 |   24 |   28 |   36 |   27 |
 
|                            66.163.77.xx -    0 |   45 |   45 |   26 |   33 |   42 |   32 |
 
|ix-3-1-0-0.tcore1.00S-Seattle.as6453.net -    0 |   45 |   45 |   69 |   73 |   93 |   72 |
 
|  if-14-2.tcore1.PDI-PaloAlto.as6453.net -    0 |   45 |   45 |   73 |   76 |   96 |   74 |
 
|   if-2-2.tcore2.PDI-PaloAlto.as6453.net -    0 |   45 |   45 |   81 |   83 |  106 |   83 |
 
|  Vlan3254.icore1.SQN-SanJose.as6453.net -   20 |   26 |   21 |   77 |   85 |   94 |   91 |
 
|                           192.205.37.65 -    0 |   45 |   45 |   81 |   86 |  135 |   83 |
 
|                    cr1.sffca.ip.att.net -   14 |   30 |   26 |   83 |   88 |  120 |   91 |
 
|                    cr1.la2ca.ip.att.net -    0 |   45 |   45 |   82 |   87 |  114 |   86 |
 
|                  gar20.la2ca.ip.att.net -    0 |   45 |   45 |   80 |   97 |  232 |  116 |
 
|               12-122-254-234.attens.net -    0 |   45 |   45 |   81 |   84 |  105 |   96 |
 
|mdf001c7613r0003-gig-12-1.lax1.attens.net -    0 |   45 |   45 |   81 |   90 |  284 |   83 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|                   No response from host -  100 |   10 |    0 |    0 |    0 |    0 |    0 |
 
|________________________________________________|______|______|______|______|______|______|
 
   WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud Provider
 

*Vlan3254.icore1.SQN-SanJose.as6453.net = 20 packet loss. (Tata Communications)

This has been ongoing for weeks now yet again. As users we have no control over this what so ever. This company is notoriously bad for packet loss issues as can be verified by google.

Could we please have someone at Shaw look into this and pass word along to them that there needs to be some improvement here? 20% packet loss pretty much destroys your connection no matter what it is capable of... anything after it doesn't have a snowballs chance in hell.


rustydusty

join:2009-09-29
Red Deer, AB

Shaw directly peers with a few backbone providers, this is definitely not one of them. Contacting them would be almost worthless as there's only so much they can do. This is also a small issue for them, mostly for the fact that this isn't affecting that many people.

Sean may chime in and pass the info along to whoever he can, but I doubt it will go anywhere. I've had similar issues and it's just the joys of routing issues. Telus is even worse for this and it happens even more often with them.


rotohoto

join:2012-03-31
canada
kudos:1

Shaw is directly connected to Tata. I don't know what the details of their arrangement are as far as peering/transit, but there is a direct connection there.

OP: Do you actually see that same loss when pinging a host on the far side of those hops? One thing I've noticed with MTR (and actually with Tata in particular) is it will sometimes make packet loss seem worse than it is, or make it seem like it's happening when it's not.

In tickets I've had with tata they are open about the fact they rate limit and deprioritze ICMP traffic directed at/generated by their routers. So I've seen instances where pinging their interfaces tells me I'm getting 50% packet loss, but pinging something going through that same interface shows 0% packet loss.



rustydusty

join:2009-09-29
Red Deer, AB
reply to Baud1200

Unless recently, they never did before. Even for direct peers, routing can have issues and Shaw doesn't pay attention to routes/packet loss unless it's on their direct network. Best bet, wait it out and let things resolve themselves. We've all had to do it before, and unfortunately it can takes months.



Baud1200

join:2003-02-10
Reviews:
·Shaw
reply to rotohoto

Up to 65% loss at 1AM.. really hope the thing melts into the ground as it seems that's the only hope of getting the hardware addressed.

What was happening was I was playing Diablo III (trying) when I started to see in game latency of over 4000. Ping was not terrible to the host itself but when i ran this trace to see where the packet loss was occurring the app fluctuates to show packet loss on this router exactly when the game lags out and becomes unplayable.



Baud1200

join:2003-02-10
Reviews:
·Shaw
reply to rustydusty

said by rustydusty:

Unless recently, they never did before. Even for direct peers, routing can have issues and Shaw doesn't pay attention to routes/packet loss unless it's on their direct network. Best bet, wait it out and let things resolve themselves. We've all had to do it before, and unfortunately it can takes months.

Thats what i was afraid of.. did some research on it before i posted here. Doesn't look good as this has been a problem router for years and various forums say to avoid this company at all costs.

Heres another user on shaw Edmonton with the same issues...
»blues.wowraider.ru/blue/us2/5633···my-WinXP


rustydusty

join:2009-09-29
Red Deer, AB
reply to Baud1200

A lot of big game server providers use cheaper backbone providers, which generally you notice higher pings, packet loss and bad routing. Obviously they are more concerned with capacity and not quality. If you look at a provider like Internap, or Level3, they are a higher end provider and rarely do I ever see issues with any host running on them.



Baud1200

join:2003-02-10
Reviews:
·Shaw

I could see if it was server lag that everyone was experiencing at the same time, or if it was the last few nodes slowing down, but i only get this latency when this router is misbehaving.

As for what you said about them handling the ICMP traffic differently that is a possibility however the fact that the throttling of ICMP kicked in where it hadn't before would indeed point to that fact that the router is at peak /overloaded.

For it to be a router at the end of the line in a cheap hosting facility, i could understand but for it to be in the middle of the Shaw and ATT network.. that seems a precarious place for a piece of equipment to be sitting producing these kind of flakey results... isn't this affecting quite a few people and most likely all the recent lag /page load issues reported on this forum?

Surely there must be some large scale business customers (or ISP's) that deal with this company that also find this unacceptable.


stolen

join:2004-04-12
Calgary, AB
kudos:2
reply to Baud1200

If you're not seeing latency past the point where you see packetloss, it only means that the router that's reporting packetloss is de-prioritizing ICMP. This is rather common. It's about as concerning as the 100% packetloss on your second hop.

Run it, and you'll see that it's likely there even when ping latency within the game is good.


stolen

join:2004-04-12
Calgary, AB
kudos:2
reply to Baud1200

Also, I've quite often seen these spikes in game and I'm not going through Tata:

                             My traceroute  [v0.80]
derf (0.0.0.0)                                         Mon Jun 25 08:30:35 2012
Keys:  Help   Display mode   Restart statistics   Order of fields   quit
                                       Packets               Pings
 Host                                Loss%   Snt   Last   Avg  Best  Wrst StDev
 1. ???
 2. 64.59.132.69                      0.0%   121   17.8  13.6   7.3  34.7   4.2
 3. 66.163.71.137                     0.0%   121   15.4  14.6   6.8  31.7   4.6
 4. 66.163.77.89                      0.0%   121   29.3  30.3  21.7  45.6   4.8
 5. 66.163.76.138                     0.0%   121   29.3  30.7  25.9  43.1   3.9
 6. xe-8-3-1.edge1.Seattle3.Level3.n  0.0%   121   26.3  30.0  24.5  90.8  10.0
 7. 192.205.36.153                    0.0%   121   29.4  33.3  24.9  88.0  11.4
 8. cr1.st6wa.ip.att.net              0.0%   121   91.4  89.2  83.6 133.5   6.0
 9. cr2.sffca.ip.att.net              0.0%   121   91.0  90.2  84.5 115.3   4.6
10. cr2.la2ca.ip.att.net              0.0%   121   91.7  89.2  83.6 109.2   4.2
11. gar29.la2ca.ip.att.net            0.0%   120   84.3  99.2  82.3 216.5  31.4
12. 12-122-254-234.attens.net         0.0%   120   83.5  90.8  82.6 275.4  21.0
    12-122-254-238.attens.net
13. mdf001c7613r0003-gig-12-1.lax1.a  0.0%   120   86.7  89.3  80.2 287.3  29.7
14. ???
 


Baud1200

join:2003-02-10
Reviews:
·Shaw

Did you take this from Calgary??
What a beautiful route through seattle from L3
(xe-8-3-1.edge1.Seattle3.Level3.n )
in one hop to the 192.xx vs from Edmonton you get this mess to the same destination...

| ix-3-1-0-0.tcore1.00S-Seattle.as6453.net
| if-14-2.tcore1.PDI-PaloAlto.as6453.net
| if-2-2.tcore2.PDI-PaloAlto.as6453.net
| Vlan3254.icore1.SQN-SanJose.as6453.net

What gives? Between seeing this and the city waking up and removing fluoride from the water supply, almost fixing to move there! Now if i could just find a location with the shaw fiber to premises testing...



rustydusty

join:2009-09-29
Red Deer, AB
reply to Baud1200

What's the destination? I'm in RD, let's see what I get. Not sure if Biz packages are different routing or not.



Baud1200

join:2003-02-10

I was hitting 12.130.244.193


stolen

join:2004-04-12
Calgary, AB
kudos:2

That might be the difference. I was using 12.129.209.68.
Lets see what I get with your IP.
Nope, pretty much the same.

      My traceroute  [v0.80]
derf (0.0.0.0)                                         Tue Jun 26 09:28:48 2012
Keys:  Help   Display mode   Restart statistics   Order of fields   quit
                                       Packets               Pings
 Host                                Loss%   Snt   Last   Avg  Best  Wrst StDev
 1. ???
 2. 64.59.132.69                      0.0%     7   14.3  13.2   8.7  17.3   3.1
 3. rc2so-pos0-8-1-0.cg.shawcable.ne  0.0%     7   21.4  13.6  10.0  21.4   4.1
 4. rc2wh-pos0-8-2-0.vc.shawcable.ne  0.0%     7   27.6  46.3  23.6  85.2  22.7
 5. rc4wt-pos1-0-0.wa.shawcable.net   0.0%     6   28.8  31.3  27.0  46.1   7.3
 6. xe-8-3-1.edge1.Seattle3.Level3.n  0.0%     6   40.0  45.8  27.9  81.0  20.9
 7. 192.205.36.153                    0.0%     6   30.0  27.2  25.3  30.0   1.6
 8. cr1.st6wa.ip.att.net              0.0%     6   87.5  85.3  82.7  88.3   2.3
 9. cr2.sffca.ip.att.net              0.0%     6   87.8  89.8  87.1  93.3   2.2
10. cr2.la2ca.ip.att.net              0.0%     6   87.9  88.8  86.6  92.7   2.1
11. gar29.la2ca.ip.att.net            0.0%     6   84.9  96.8  83.8 156.5  29.2
12. 12-122-254-238.attens.net         0.0%     6   85.3  85.5  82.1  87.9   2.1
    12-122-254-234.attens.net
13. mdf001c7613r0004-gig-10-1.lax1.a  0.0%     6   85.2  85.4  83.7  88.8   1.8
14. ???
 
 
And yes, from Calgary.


Baud1200

join:2003-02-10
Reviews:
·Shaw

|                       Host             -%Loss  | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
|                DarkHorse.PrisonProvince -    0 |   25 |   25 |    0 |    0 |    0 |    0 |
|                   No response from host -  100 |    5 |    0 |    0 |    0 |    0 |    0 |
|                           64.59.xxx.xxx -    0 |   25 |   25 |    9 |   16 |   24 |   18 |
|                            66.163.78.xx -    0 |   25 |   25 |   25 |   29 |   37 |   26 |
|         rc3wt-ge14-0-0.wa.shawcable.net -    0 |   25 |   25 |   26 |   30 |   43 |   30 |
|ix-0-0-2-0.tcore1.00S-Seattle.as6453.net -    0 |   25 |   25 |   71 |   75 |   83 |   73 |
|  if-14-2.tcore1.PDI-PaloAlto.as6453.net -    0 |   25 |   25 |   72 |   76 |   85 |   72 |
|   if-2-2.tcore2.PDI-PaloAlto.as6453.net -    0 |   25 |   25 |   73 |   76 |   84 |   73 |
|  Vlan3254.icore1.SQN-SanJose.as6453.net -   24 |   13 |   10 |    0 |   77 |   85 |   73 |
|                           192.205.37.65 -    0 |   25 |   25 |   73 |   84 |  138 |   76 |
|                    cr1.sffca.ip.att.net -   12 |   17 |   15 |   81 |   86 |   98 |   81 |
|                    cr1.la2ca.ip.att.net -    0 |   25 |   25 |   84 |   88 |   98 |   87 |
|                  gar20.la2ca.ip.att.net -    4 |   25 |   24 |   82 |   99 |  234 |   84 |
|               12-122-254-234.attens.net -    0 |   25 |   25 |   81 |   95 |  234 |   82 |
|mdf001c7613r0004-gig-10-1.lax1.attens.net -   0 |   25 |   25 |   83 |   89 |  139 |   84 |
|                   No response from host -  100 |    5 |    0 |    0 |    0 |    0 |    0 |
|________________________________________________|______|______|______|______|______|______|
 

Thats to the address you provided, taken just a few minutes ago... looks like Edmonton gets the shaft here with a direct route directly through the worse of the 2 providers. Also note its lagging out again...

ravenchilde

join:2011-04-01
kudos:2

said by Baud1200:

Thats to the address you provided, taken just a few minutes ago... looks like Edmonton gets the shaft here with a direct route directly through the worse of the 2 providers. Also note its lagging out again...

As Stolen explained, it may not be lagging out, it may just have ICMP as its lowest processing priority. This isn't uncommon on internet backbone devices. As the next hop is packet loss free, its not actually dropping your packets.


Baud1200

join:2003-02-10
Reviews:
·Shaw

4 edits

By lagging out i mean the game hitching, to 3K ping again so i kill it and then see this result in tracert. Again, when not getting lag in the game tracert through this router is ok, its a direct correlation of the 2...

Also notice that everything (ATT routes) after your route through L3 was clean in terms of packet loss, yet mine going through ATT was dirty as heck due to the issues of that lossy router before it no doubt. This also shows in the peak worst ping category where you can clearly see its spikes to almost 3x that of yours.

That's my issue in a nutshell basically.. anything past that route is essentially "broked"