New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
I've also adjusted Maxis.com.my / AS9534 to see if that helps. Maxis users should see a marked improvement once NTT comes up as well (hopefully next week).
I think right now Maxis itself have network problem.
Are you sure Switchnap is faster? Looks like it's about equal to me:
ping 202.151.252.42
Sending 1, 16-byte ICMP Echo to 202.151.252.42, timeout 5000 msec, TTL 64
Type Control-c to abort
Reply from 202.151.252.42 : bytes=16 time=178ms TTL=249
Success rate is 100 percent (1/1), round-trip min/avg/max=178/178/178 ms.
Dear Rob,
Possible to peer with celcom.com.my too? It has the 2nd largest userbase in Malaysia after tmnet.
TQ, I'm on Maxis
but testing just now, ping certainly got better below 300 but speed still throttled.
I'll just wait and see. already got a backup vpn
You can test download and upload speeds.
http://speedtest.switchnap.com/
http://speedtest.switchnap.com/speedtest/random4000x4000.jpg
./speedtest_cli.py --server 1431
just got access to my friend unifi TMnet
here's the mtr
Packets Pings
Host Loss% Snt Last Avg Best Wrst StDev
1. node21.buyvm.net 0.0% 31 0.1 0.1 0.1 0.3 0.0
2. 10.1.1.5 0.0% 31 0.5 0.7 0.4 3.4 0.6
3. te3-4.core1.las1.fiberhub.net 0.0% 31 0.5 1.1 0.4 9.6 1.7
4. 10ge3-2.core1.las1.he.net 0.0% 31 0.5 3.4 0.4 12.7 4.5
5. 10ge15-4.core1.lax2.he.net 0.0% 31 7.9 13.3 7.8 53.7 8.8
6. 10ge2-1.core1.lax1.he.net 0.0% 31 18.3 10.9 7.9 18.3 3.6
7. LAX-GE1.tm.net.my 0.0% 30 8.6 15.6 8.4 106.5 19.8
8. 10.55.36.76 0.0% 30 262.8 269.8 262.7 328.6 15.3
10.55.36.106
10.55.36.40
10.55.36.104
9. 175.144.88.70 0.0% 30 266.0 269.2 265.0 287.1 5.4
175.144.88.58
115.132.109.230
10. 175.139.90.104 0.0% 30 268.6 270.0 266.1 288.0 4.2
but suddenly my maxis speed back to normal
mtr to my ip Maxis
Packets Pings
Host Loss% Snt Last Avg Best Wrst StDev
1. node21.buyvm.net 0.0% 69 0.3 0.4 0.3 0.7 0.1
2. 10.1.1.5 0.0% 69 0.9 2.3 0.5 10.7 2.9
3. te3-4.core1.las1.fiberhub.net 0.0% 69 0.9 1.0 0.6 7.5 1.0
4. 10ge3-2.core1.las1.he.net 0.0% 69 4.8 3.4 0.6 11.6 3.8
5. 10ge15-4.core1.lax2.he.net 0.0% 69 8.1 10.0 8.0 20.0 3.3
6. las-b3-link.telia.net 1.4% 69 8.2 9.2 8.0 58.3 6.1
7. las-bb1-link.telia.net 2.9% 69 15.1 14.8 8.6 53.9 6.1
8. ae8.edge1.LosAngeles.Level3.net 0.0% 69 8.7 9.1 8.4 31.2 2.8
9. ae-4-90.edge1.LosAngeles6.Level3 0.0% 69 8.5 8.6 8.3 10.2 0.3
10. MAXIS-BROAD.edge1.LosAngeles6.Le 0.0% 69 9.0 9.9 8.7 35.0 3.9
11. 202.151.252.253 1.5% 69 198.6 199.1 198.0 219.6 2.9
12. 202.151.241.129 2.9% 69 197.9 197.9 197.4 200.0 0.4
13. 202.151.241.134 0.0% 68 201.1 201.3 200.9 203.3 0.3
14. bun-eth1.kpgp-pe60e.maxis.net.my 0.0% 68 199.6 200.1 197.6 216.8 2.8
15. ???
*edit
haha just a blip i guess
back to throttled speed. but the latency definitely way better than before! now can get to us under 100ms
update still no improvement
BTW this is my friend on PI Wimax connection
mtr from my vps
Packets Pings
Host Loss% Snt Last Avg Best Wrst StDev
1. node21.buyvm.net 6.2% 275 0.4 0.4 0.0 10.1 0.7
2. 10.1.1.5 0.0% 275 0.7 1.6 0.3 11.6 2.3
3. 38.88.14.9 10.2% 275 1.0 19.2 0.6 393.6 49.5
4. te2-4.ccr01.las02.atlas.cogentco 11.3% 275 1.1 10.8 0.9 204.8 34.6
5. te0-4-0-5.mpd21.lax01.atlas.coge 0.0% 275 8.3 8.8 8.1 22.2 1.2
6. be2021.ccr21.lax04.atlas.cogentc 0.0% 275 8.8 8.8 8.1 18.1 0.9
7. ntt.lax04.atlas.cogentco.com 0.0% 275 8.8 9.0 8.3 17.3 0.8
8. ae-2.r20.lsanca03.us.bb.gin.ntt. 0.0% 275 8.5 12.5 8.0 100.2 10.8
9. ae-6.r20.osakjp02.jp.bb.gin.ntt. 1.8% 275 120.1 119.2 112.6 150.4 6.4
10. ae-7.r22.osakjp01.jp.bb.gin.ntt. 0.0% 275 164.3 141.5 113.1 233.9 22.4
11. as-0.r20.sngpsi05.sg.bb.gin.ntt. 0.0% 274 202.3 203.4 197.9 238.7 6.5
12. ae-0.r20.sngpsi02.sg.bb.gin.ntt. 0.0% 274 188.0 187.7 182.8 220.5 5.2
13. as-0.r03.kslrml01.my.bb.gin.ntt. 0.0% 274 191.7 194.2 188.0 284.3 12.0
14. ge-0-1-0.r03.kslrml01.ce.bb.gin. 0.0% 274 201.9 199.4 193.6 389.7 14.9
15. te3-1.pkh002.p1.com.my 0.0% 274 196.0 198.1 192.5 356.7 14.1
16. te3-1.pkh022.p1.com.my 61.2% 274 193.9 194.4 190.5 217.3 3.6
17. 122.255.112.232 95.2% 274 216.2 212.4 209.0 216.8 2.9
18. ???
19. 120.140.50.73 0.4% 274 267.6 354.0 243.1 612.1 68.1
no throttle whatsoever
from my other TX vps
Packets Pings
Host Loss% Snt Last Avg Best Wrst StDev
1. 198.23.249.11 0.0% 183 0.0 0.0 0.0 0.1 0.0
2. host.colocrossing.com 0.5% 183 2.6 8.7 0.5 62.0 12.7
3. colocrossing-quadranet.quadranet 1.1% 183 1.1 11.1 0.3 188.7 22.8
4. 69.31.54.193 0.5% 183 0.5 7.6 0.3 54.4 12.3
5. as2914.xe-0-0-0.cr1.dfw1.us.nlay 1.1% 183 1.3 8.6 0.7 91.4 13.1
6. ae-1.r21.dllstx09.us.bb.gin.ntt. 0.5% 183 38.3 12.9 0.5 94.0 19.0
7. ae-4.r21.snjsca04.us.bb.gin.ntt. 0.5% 183 44.8 57.9 43.6 127.1 16.2
8. ae-6.r21.osakjp02.jp.bb.gin.ntt. 1.1% 183 164.7 174.5 154.2 257.4 17.5
9. ae-8.r23.osakjp01.jp.bb.gin.ntt. 1.6% 183 152.1 172.2 140.8 267.7 24.6
10. p64-2-0-0.r02.kslrml01.my.bb.gin 0.0% 183 242.5 247.5 231.1 288.9 12.2
11. xe-0-1-0.r02.kslrml01.my.ce.gin. 0.0% 183 240.0 249.5 230.0 419.0 22.8
12. 202.151.241.177 0.5% 183 251.3 258.6 241.9 323.7 15.0
13. 202.151.241.145 1.6% 183 241.3 244.6 230.2 286.5 11.5
14. 202.151.241.141 1.1% 183 251.5 256.6 241.3 307.5 13.4
15. bun-eth1.kpgp-pe60e.maxis.net.my 0.0% 183 282.0 246.4 228.4 295.1 14.1
16. ???
I still get the feeling that the issue is on your end.
With that being said! We'll have NTT in our own bandwidth this week hopefully, so you'll go right from your ISP to us on NTT.
Francisco
Haha after testing with my friend connection I think the same too.
glad to hear you're getting ntt.
I did my best to not put any blame on your provider but it does smell that way.
As I said, though, NTT will come in and it should help remove the potential cogent headache
Thanks for being so patient,
Francisco
Fran, can you check mine? I canceled all your vps because of that results I had always a bad connection and I know I'm not your client, but it may help the other ones
Lets wait for NTT and our cogent routes lessened and then I'll ask you to test again, deal?
Francisco
@francisco ok
Thanks
Francisco
IP: 205.185.112.31
Date/Time: 1/14/2014 9:41:33 AM
1 32 ms [115.132.110.58]
2 40 ms [115.132.110.57]
3 269 ms [10.55.204.9]
4 295 ms 30gigabitethernet1-3.core1.ams1.he.net [195.69.145.150]
5 304 ms 100ge9-1.core1.lon2.he.net [72.52.92.213]
6 296 ms 100ge1-1.core1.nyc4.he.net [72.52.92.166]
7 299 ms 100ge7-2.core1.chi1.he.net [184.105.223.161]
8 299 ms 10ge3-2.core1.den1.he.net [184.105.213.86]
9 301 ms 10ge4-2.core1.slc1.he.net [184.105.222.154]
10 309 ms 10ge1-1.core1.las1.he.net [184.105.222.157]
11 307 ms vegasnap-llc.10gigabitethernet5-1.core1.lax2.he.net [64.62.249.90]
12 301 ms te1-1.core2.las1.fiberhub.net [199.47.208.10]
13 301 ms [10.1.1.6]
14 301 ms mirrors.buyvm.net [205.185.112.31]
Ping statistics for speedtest.buyvm.net
Packets: Sent = 1, Received = 1, Lost = 0 (0.0%)
Round Trip Times: Minimum = 301ms, Maximum = 301ms, Average = 301ms
From Malaysia obviously the route is not good at all
When did everyone in Malaysia buy VPS servers?
As I said, lets see how NTT helps. It should help a lot.
I do know @Rob_T pushed certain things over HE.
Francisco
Traceroute and pings will only give you minimal information on which to speculate on. Better you setup an iperf udp flow this will give you a better idea of how stable the route is (jitter and thus out of order packets) and how much loss occurs. With 300ms rtt latency you only need minimal packet loss for tcp performance to suffer, and I suspect this is what you are seeing, tcp wrongly assumes the packet loss scales with throughput and thus its pretty sucky situation. It may even be that this packet loss is happening within your ISPs network but you only notice it against buyvm because when you hit youtube, google, bing, gmail, vimeo, sourceforge etc.. the larger assets are on edge/cdn networks (I suspect MY has plenty but if not next door in SG) and the effect is massively dampened due to smaller rtt. Local forums in asia are rife with speculation that it's the ISPs deliberately doing this, and its those nasty transparent proxies*, or they deliberately throttle "international bandwidth" when the reality is that a small degree of packet loss has to be expected, even tier 1 network say (or used to say) they SLA to < 1% packet loss (on a 300ms rtt that's gonna hurt any http downloads).
* transparent proxies still suck but they [wrongly] get the blame for everything
It's been awhile ... we're addict man!!
anyway i had been your client since 2010 (on and off) ... back then it was Frantech instead.
LOL
how do I interpret this?
im using sample argument from this http://goo.gl/nQbc2Y
@Francisco While using Maxis, it's usually via Cogent to your Vegas location.
I'm afraid it will not. Not anymore.
mhm
UPC Austria (250/10):
Compared to us.edis.at:
And, for comparsion, Leaseweb NL:
Not only Asia can into bad routing.... UPC has 100million customers in Europe...