Log for
電訊台
Traceroute is the golden standard for diagnose network issues
onr of the hop is 70ms i remember
but this makes replication slow...
traceroute to 47.254.2xx.16x (47.254.2xx.16x), 30 hops max, 60 byte packets
1 * * *
2 117.121.248.216 (117.121.248.216) 0.753 ms 0.852 ms 0.779 ms
3 timedc.lag50.fr3.sin.llnw.net (111.119.23.5) 0.728 ms 0.732 ms 1.111 ms
4 * * *
5 ae2-er-01-csfcb.ni.time.net.my (223.28.26.86) 8.743 ms 8.760 ms 8.727 ms
6 203.121.113.214 (203.121.113.214) 87.884 ms 87.334 ms 87.937 ms
7 * * *
8 * * *
9 * * *
10 * * *
11 * * *
12 * * *
13 * * *
14 * * *
15 * * *
16 * * *
17 * * *
18 * * *
19 * * *
20 * * *
21 * * *
22 * * *
23 * * *
24 * * *
25 * * *
26 * * *
27 * * *
28 * * *
29 * * *
30 * * *
see
Check power on is the golden standard
The distance between Kuala Lumpur and Singapore is 316 km
speed of light = 0.00333564095 millisecond per kilometer
then theoretical max latency = 316 * 0.00333564095 = 1 ms..
啲router switch唔洗process time?未計佢係19
U have money u go better lor
Convergence is needed, otherwise... U see fucked up
ah, so convergence is an optimization technique?
like dijkstra flow combine step
Simplify and save energy to serve jerk like u lol
im not doing anything illegal but just running disk replication
this malay instance is the bottleneck apparently
Wait and see lor. Nothing to do for such... Hmm you can pay more lor
時代冰室個大媽無咩禮貌,坐左好耐都落唔到單,佢話「等多陣得唔得」
Traceroute will give u some information
theory first, practice second
traceroute to 193.123.x4x.84 (193.123.x4x.84), 30 hops max, 60 byte packets
1 * * *
2 117.121.248.216 (117.121.248.216) 0.499 ms 0.561 ms 0.546 ms
3 63.218.229.9 (63.218.229.9) 1.045 ms 1.060 ms 1.045 ms
4 TenGE0-0-0-2.br03.sin02.pccwbtn.net (63.218.228.138) 0.985 ms TenGE0-2-0-15.br03.sin02.pccwbtn.net (63.218.228.182) 2.262 ms HundredGE0-3-0-0.br03.sin02.pccwbtn.net (63.223.34.134) 0.905 ms
5 TenGE0-0-0-1.br03.sin02.pccwbtn.net (63.218.228.110) 0.932 ms 1.185 ms HundredGE0-5-0-0.br03.sin02.pccwbtn.net (63.223.34.142) 0.788 ms
6 63-218-165-50.static.pccwglobal.net (63.218.165.50) 167.062 ms 167.334 ms 167.087 ms
7 if-ae-20-2.tcore2.svw-singapore.as6453.net (180.87.96.22) 209.157 ms if-ae-32-4.tcore2.hk2-hongkong.as6453.net (116.0.93.168) 207.085 ms if-ae-32-2.tcore2.hk2-hongkong.as6453.net (180.87.15.81) 193.285 ms
8 if-ae-2-2.thar1.hk2-hongkong.as6453.net (116.0.82.62) 193.231 ms 193.169 ms if-ae-37-4.tcore1.hk2-hongkong.as6453.net (116.0.93.147) 194.809 ms
9 if-ae-37-2.tcore1.hk2-hongkong.as6453.net (116.0.93.145) 191.437 ms if-ae-37-6.tcore1.hk2-hongkong.as6453.net (116.0.93.137) 193.075 ms if-ae-2-2.thar1.hk2-hongkong.as6453.net (116.0.82.62) 191.275 ms
10 if-et-5-2.hcore1.aseoa-seoul.as6453.net (116.0.67.38) 206.644 ms 201.547 ms 201.525 ms
11 140.91.214.9 (140.91.214.9) 95.390 ms 140.91.214.15 (140.91.214.15) 101.014 ms 140.91.214.9 (140.91.214.9) 96.039 ms
12 193.123.242.84 (193.123.x4x.84) 95.995 ms 95.945 ms 95.927 ms
what the
They got the route, and thats it
呢個問題同啲人問點解架飛機唔直線飛去某地方,差唔多lv
新加坡 - 香港 - 韓國
新加坡 - 香港 - 台灣 - 韓國
So, go to complain and complain la
我一直以爲latency就係triangular inequality咁簡單....
尋日先同你講完 BGP best path algorithm
i have goldfish memory...
其實純粹考慮 administrative distance
Bandwidth 同 latency 其實個 algorithm 冇考慮,係 infer 出嚟
所以成日要用 BGP attributes 人手微調
佢啲計數係得意嘅,好似佢嘅下個node就到韓國咁,一條dark fiber直到
就算係 dark fibre 可能都要係香港 patch 一次
之前 WTT 未比 HKBN 買嘅時候有一條咁嘅嘢
traceroute to 199.2x1.1x8.18 (199.2x1.1x8.18), 30 hops max, 60 byte packets
1 * * *
2 11.221.217.25 (11.221.217.25) 0.162 ms 11.221.218.29 (11.221.218.29) 0.160 ms 11.221.217.29 (11.221.217.29) 0.154 ms
3 11.60.202.109 (11.60.202.109) 4.858 ms 11.60.202.105 (11.60.202.105) 5.081 ms 5.201 ms
4 11.52.55.186 (11.52.55.186) 0.426 ms 11.52.55.174 (11.52.55.174) 0.292 ms 11.52.55.186 (11.52.55.186) 0.390 ms
5 116.251.73.190 (116.251.73.190) 0.335 ms * *
6 xe-0-0-5.r20.kslrml02.my.bb.gin.ntt.net (203.78.193.53) 1.353 ms xe-0-0-2.r21.kslrml02.my.bb.gin.ntt.net (203.78.193.49) 0.406 ms 0.400 ms
7 ae-8.r21.sngpsi07.sg.bb.gin.ntt.net (129.250.2.168) 6.751 ms ae-1.r21.kslrml02.my.bb.gin.ntt.net (129.250.7.43) 3.552 ms ae-8.r21.sngpsi07.sg.bb.gin.ntt.net (129.250.2.168) 6.725 ms
8 ae-8.r21.sngpsi07.sg.bb.gin.ntt.net (129.250.2.168) 6.691 ms ae-17.r31.tokyjp05.jp.bb.gin.ntt.net (129.250.2.243) 70.040 ms ae-8.r21.sngpsi07.sg.bb.gin.ntt.net (129.250.2.168) 8.857 ms
9 ae-3.r01.tokyjp08.jp.bb.gin.ntt.net (129.250.6.133) 72.491 ms ae-17.r31.tokyjp05.jp.bb.gin.ntt.net (129.250.2.243) 73.768 ms 73.905 ms
10 ae-3.r00.tokyjp08.jp.bb.gin.ntt.net (129.250.6.129) 73.850 ms 61.120.144.38 (61.120.144.38) 84.970 ms ae-3.r00.tokyjp08.jp.bb.gin.ntt.net (129.250.6.129) 71.691 ms
11 tge12-1.fr4.tyo3.llnw.net (111.119.25.124) 83.274 ms 61.120.144.30 (61.120.144.30) 95.306 ms tge12-1.fr4.tyo3.llnw.net (111.119.25.124) 98.086 ms
12 * ve5.fr4.tyo3.llnw.net (203.77.184.230) 78.147 ms tge12-1.fr3.tyo3.llnw.net (111.119.25.126) 86.470 ms
13 107-155-95-253.static.hvvc.us (107.155.95.253) 109.006 ms lag42.fr4.sin.llnw.net (203.77.185.11) 80.498 ms lag58.fr4.sin.llnw.net (111.119.23.32) 95.175 ms
14 * 107-155-95-253.static.hvvc.us (107.155.95.253) 108.264 ms 96.621 ms
15 199.2x1.1x8.18 (199.241.138.18) 75.091 ms 81.470 ms 74.761 ms
哇 從馬來去新加坡回程經日本真係過份
因為你個 target 冇買 ntt transit
You minimise the administrative distance by more peering and transit
And not all transit works nice with your intended geographical area