Rate limiting in effect?

75 views
Skip to first unread message

Mark van der Sanden

unread,
Mar 29, 2012, 5:26:56 AM3/29/12
to MailChimp API Discuss
Hi! I have a problem with a client of mine. Most of the time I get a
timeout when trying to connect to the MailChimp API, but not always.

Now I was wondering .. Maybe you have some kind of rate limiting on an
IP level. So if other sites on the same machine flood your API, then
my client's site will not be able to connect.

See a little test on http://sandbox.magentics.nl/test/mailchimp/mcapi_lists.php

The originating IP is 46.21.225.7

Is there something I can do to make this work?

Thanks!

jesse

unread,
Mar 29, 2012, 5:28:57 AM3/29/12
to mailchimp-...@googlegroups.com

Mark van der Sanden

unread,
Mar 29, 2012, 6:34:38 AM3/29/12
to mailchimp-...@googlegroups.com
Jesse,

Thanks for your reply! I had already read those, but couldn't find any useful answer in it.

It's just a strange problem. Only this machine seems to have problems connecting to the API with this specific API key.
Other machines have no problem and other API keys also have no problems at all at the same machine.

Regards,
Mark

Mark van der Sanden

unread,
Mar 29, 2012, 6:44:51 AM3/29/12
to mailchimp-...@googlegroups.com
What I forgot to mention, is that the other API key that I tried at the same machine uses another hostname to connect to (the problematic key ends in 'us4' while the working key ends in 'us1').


On Thursday, March 29, 2012 11:28:57 AM UTC+2, jesse wrote:

jesse

unread,
Mar 29, 2012, 6:57:19 AM3/29/12
to mailchimp-...@googlegroups.com
Then you should also be familiar with how the API endpoints work:

http://apidocs.mailchimp.com/api/rtfm/

You have some sort of connectivity issue you'll need to troubleshoot.


jesse

Mark van der Sanden

unread,
Mar 29, 2012, 9:38:56 AM3/29/12
to mailchimp-...@googlegroups.com
I can ping us4.api.mailchimp.com with no problem at all:

PING us4.api.mailchimp.com (184.173.96.194) 56(84) bytes of data.
64 bytes from 184.173.96.194-static.reverse.softlayer.com (184.173.96.194): icmp_seq=1 ttl=46 time=165 ms
64 bytes from 184.173.96.194-static.reverse.softlayer.com (184.173.96.194): icmp_seq=2 ttl=46 time=165 ms
64 bytes from 184.173.96.194-static.reverse.softlayer.com (184.173.96.194): icmp_seq=3 ttl=46 time=165 ms
64 bytes from 184.173.96.194-static.reverse.softlayer.com (184.173.96.194): icmp_seq=4 ttl=46 time=165 ms

But connecting to port 80 or 443 is not possible from this machine, but I *can* connect to us1 and us2.

Is there anything else I can do to do some more testing?

Regards,
Mark

Mark van der Sanden

unread,
Mar 29, 2012, 10:10:27 AM3/29/12
to mailchimp-...@googlegroups.com
And can you please check if 46.21.225.7 is somehow blacklisted?


On Thursday, March 29, 2012 12:57:19 PM UTC+2, jesse wrote:

jesse

unread,
Mar 29, 2012, 10:25:04 AM3/29/12
to mailchimp-...@googlegroups.com
That IP is not blocked. You can use something like tcptraceroute to see where you're being blocked.

# tcptraceroute -P80 us4.api.mailchimp.com


jesse

Mark van der Sanden

unread,
Mar 30, 2012, 5:39:53 AM3/30/12
to mailchimp-...@googlegroups.com
I got an answer of the sysadmin:

traceroute to us4.api.mailchimp.com (184.173.96.194), 30 hops max, 40 byte packets
1 lb1.c101.internal (10.2.1.24) 0.522 ms 0.499 ms 0.488 ms
2 ip-46-21-224-1.ams.byte.nl (46.21.224.1) 0.745 ms 0.851 ms 0.845 ms
3 0.ge-0-1-0.xr3.3d12.xs4all.net (194.109.12.1) 14.851 ms 14.953 ms 14.947 ms
4 asd2-rou-1002.NL.eurorings.net (134.222.97.17) 0.736 ms 0.731 ms 0.722 ms
5 asd2-rou-1002.NL.eurorings.net (134.222.230.142) 47.248 ms 47.327 ms 47.481 ms
6 adm-b5-link.telia.net (213.155.129.21) 1.094 ms 0.737 ms 0.725 ms
7 adm-bb1-link.telia.net (213.155.134.42) 1.029 ms adm-bb2-link.telia.net (80.91.253.170) 1.051 ms adm-bb2-link.telia.net (80.91.246.100) 22.630 ms
8 ldn-bb1-link.telia.net (80.91.245.106) 8.124 ms 8.128 ms 8.117 ms
9 ash-bb1-link.telia.net (80.91.246.68) 83.610 ms 83.589 ms 83.584 ms
10 xe-3-0-1.bbr01.eq01.wdc02.networklayer.com (213.248.78.130) 100.255 ms 100.253 ms 99.317 ms
11 ae0.bbr01.tl01.atl01.networklayer.com (173.192.18.153) 99.848 ms 99.834 ms 99.598 ms
12 * * *
13 ae5.dar02.sr01.dal05.networklayer.com (173.192.18.217) 165.800 ms ae6.dar01.sr01.dal05.networklayer.com (50.97.18.195) 165.795 ms ae5.dar02.sr01.dal05.networklayer.com (173.192.18.217) 165.898 ms
14 po1.fcr02.sr02.dal05.networklayer.com (173.192.118.139) 166.840 ms po2.fcr02.sr02.dal05.networklayer.com (173.192.118.141) 173.601 ms 173.692 ms
15 * * *
16 * * *
17 * * *
18 * * *
19 *^C

If the sysadmin does that on a different server in the same network, it adds another line (the IP of us4):
14 * 184.173.96.194-static.reverse.softlayer.com (184.173.96.194) 165.292 ms *

IMHO the traffic gets blocked on your side.
As I mentioned earlier, it's not 100% of the time. Sometimes, say one in 30 times I try, I got no problem and can make a connection on port 80.

Thanks!

Mark

jesse

unread,
Mar 30, 2012, 6:05:01 AM3/30/12
to mailchimp-...@googlegroups.com
A traceroute is not checking either port 80 or 443, which is why I offered the tcptraceroute suggestion. I also have a remote server which doesn't properly completely traceroutes, so that doesn't surprise me in the least.

If you do a valid test and still stop at 173.192.118.139, you can let us know. But, that's not our equipment, so if there's actually some sort of block there it's likely automated and due to your server doing bad stuff, so you'd also want to look into what it's really doing.


jesse

Mark van der Sanden

unread,
Mar 30, 2012, 6:40:51 AM3/30/12
to mailchimp-...@googlegroups.com
It *is* a tcptraceroute on port 80. You are right that regular traceroutes don't complete properly (the addition that it completes properly on another machine is enough proof of having executed the right command).

Mark
Reply all
Reply to author
Forward
0 new messages