Route from australia to 8.8.8.8 is broken

[pa...@cernik.name]

Hello,

Approximately since 2018-10-05 I noticed that latency of 8.8.8.8 responses got significantly slower. I tested the route and got following output:

C:\temp>tracert -d 8.8.4.4

Tracing route to 8.8.4.4 over a maximum of 30 hops

  1    <1 ms    <1 ms    <1 ms  10.100.1.254

  2    13 ms     6 ms     4 ms  45.248.50.1

  3     4 ms     4 ms     5 ms  45.248.48.1

  4     7 ms     5 ms     4 ms  120.29.243.153

  5    29 ms    50 ms    32 ms  172.17.46.133

  6    29 ms    28 ms    29 ms  103.26.68.56

  7    31 ms    31 ms    32 ms  108.170.247.49

  8    30 ms    30 ms    31 ms  72.14.234.31

  9    32 ms    31 ms    29 ms  8.8.4.4

Trace complete.

C:\temp>tracert -d 8.8.8.8

Tracing route to 8.8.8.8 over a maximum of 30 hops

  1     1 ms    <1 ms    <1 ms  10.100.1.254

  2     5 ms     5 ms     5 ms  45.248.50.1

  3    12 ms    13 ms    12 ms  45.248.48.1

  4     5 ms     5 ms     4 ms  120.29.243.153

  5    32 ms    29 ms    30 ms  172.17.46.133

  6    30 ms    30 ms    32 ms  103.26.68.56

  7    30 ms    30 ms    30 ms  108.170.247.42

  8   127 ms   155 ms   124 ms  108.170.234.70

  9   131 ms   120 ms   121 ms  108.170.225.101

 10   123 ms   122 ms   121 ms  216.239.35.159

 11     *        *        *     Request timed out.

 12     *        *        *     Request timed out.

 13     *        *        *     Request timed out.

 14     *        *        *     Request timed out.

 15     *        *        *     Request timed out.

 16     *        *        *     Request timed out.

 17     *        *        *     Request timed out.

 18     *        *        *     Request timed out.

 19     *        *        *     Request timed out.

 20   124 ms   120 ms   120 ms  8.8.8.8

You can see that 8.8.4.4 seems fine. However 8.8.8.8 is acting weird - after it gets to AS15169 (google's network) it hops many times and instead of typical 26-30ms it takes 100ms more.

I tested this on multiple ISPs in Australia and it is happening everywhere.

I dont think anyone except google could possibly affect routing within AS15169, therefore I believe you guys should be aware of this situation so you can fix it.

Thanks for any reply

Pavel Cernik

[Puneet Sood]

Hello,

Thanks for reporting the problem with 8.8.8.8.

We are experiencing an overload in Australia which has resulted in draining 8.8.8.8 from our serving locations in Australia. This is a temporary problem and we expect to have it fixed in the next couple of weeks.

-Puneet

[Alex Dupuy]

In general, due to the imbalance between queries to 8.8.8.8 and 8.8.4.4, you will pretty much never get worse service from 8.8.4.4 than you get from 8.8.8.8 and it can be significantly better in some circumstances.

If your resolver pays attention to ordering (and doesn't balance queries based on latency) you should put 8.8.4.4 first.