As part of a project I've written a very simplistic DNS server whose only purpose is to resolve queries for the zone it serves, and to store the IP addresses of the server that made the query.
I've noticed that if I use dig, my DNS server gets queried multiple times - sometimes from the same IP address. Why does this happen? Is it due to the unreliable nature of UDP?
For example, here's a dig reply I made:
C:\Data>dig xyz.dns.example.com
; <<>> DiG 9.10.4-P2 <<>> xyz.dns.example.com
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 2539
;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
;; QUESTION SECTION:
;xyz.dns.example.com. IN A
;; ANSWER SECTION:
xyz.dns.example.com. 12321 IN A 50.16.166.175
;; Query time: 224 msec
;; SERVER: 192.168.1.1#53(192.168.1.1)
;; WHEN: Thu Aug 11 15:07:42 Eastern Daylight Time 2016
;; MSG SIZE rcvd: 77
In this example, the zone file for example.com has an NS record for dns.example.com which is where my simplistic DNS server runs. Fror this one query, my server was called 4 times from 2 different IP addresses.
I also noticed that I'm supposedly returning an "Additional" record, but the data I return in bytes 10 and 11 are clearly 0. Could this be causing a problem?
Try dig's +trace option:
dig example.com +trace
Related
I'm trying to set up DNS receiving and tracking records for Mailgun. The DNS sending records were verified and work fine, but for some reason the receiving records and tracking records don't get verified.
The domain I'm using is mg.optimizeprice.com and the DNS provider I'm using is Dynadot. There are two mx records that it wants me to set up. I took screenshots of the Mailgun DNS records page as well as the Dynadot DNS records page as I have them set up right now. What do I need to change to get this to work?
Also, here is the output of dig optimizeprice.com mx:
; <<>> DiG 9.11.3-1ubuntu1.8-Ubuntu <<>> optimizeprice.com mx
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 8708
;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 1, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 512
;; QUESTION SECTION:
;optimizeprice.com. IN MX
;; ANSWER SECTION:
optimizeprice.com. 10800 IN CNAME curved-aardvark-xbc5jxmn5ja3pdaq403yoxab.herokudns.com.
;; AUTHORITY SECTION:
herokudns.com. 10 IN SOA dns1.p05.nsone.net. hostmaster.nsone.net. 1563353642 600 900 1209600 10
;; Query time: 41 msec
;; SERVER: 75.75.75.75#53(75.75.75.75)
;; WHEN: Wed Jul 17 01:55:08 PDT 2019
;; MSG SIZE rcvd: 176
We are doing API calls to Docusign, which fail occasionally with "getaddrinfo: Name or service not known" errors. Investigating further, we see that when we connect, name resolution fails sometimes, but only from our West datacenter location. Seems the GLB DNS for the US West can take a very long time to resolve, causing DNS client timeouts when it takes >10s to look up the address.
$ dig #1.1.1.1 www.docusign.net
; <<>> DiG 9.9.5-3ubuntu0.8-Ubuntu <<>> #1.1.1.1 www.docusign.net
; (1 server found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 49468
;; flags: qr rd ra; QUERY: 1, ANSWER: 3, AUTHORITY: 0, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 1452
;; QUESTION SECTION:
;www.docusign.net. IN A
;; ANSWER SECTION:
www.docusign.net. 22 IN CNAME www-geo.docusign.net.akadns.net.
www-geo.docusign.net.akadns.net. 22 IN CNAME www-west.docusign.net.akadns.net.
www-west.docusign.net.akadns.net. 22 IN A 162.248.184.27
;; Query time: 1 msec
;; SERVER: 1.1.1.1#53(1.1.1.1)
;; WHEN: Thu Oct 04 13:16:43 EDT 2018
;; MSG SIZE rcvd: 126
Above is a good result, which took 1msec (cached)
$ dig #1.1.1.1 www.docusign.net
; <<>> DiG 9.9.5-3ubuntu0.8-Ubuntu <<>> #1.1.1.1 www.docusign.net
; (1 server found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 21193
;; flags: qr rd ra; QUERY: 1, ANSWER: 3, AUTHORITY: 0, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 1452
;; QUESTION SECTION:
;www.docusign.net. IN A
;; ANSWER SECTION:
www.docusign.net. 6 IN CNAME www-geo.docusign.net.akadns.net.
www-geo.docusign.net.akadns.net. 6 IN CNAME www-west.docusign.net.akadns.net.
www-west.docusign.net.akadns.net. 6 IN A 162.248.184.27
;; Query time: 2725 msec
;; SERVER: 1.1.1.1#53(1.1.1.1)
;; WHEN: Thu Oct 04 13:21:29 EDT 2018
;; MSG SIZE rcvd: 126
This one is worse, as it took nearly 3s. During testing we've seen this go over 12s which will time out a lot of DNS clients and requesting apps.
Since the TTL is set to 30s, that means that every 30 seconds we have a chance at getting a timeout, our app generating errors, then a DNS success results in resumption of service. Unfortunately, this shows up as an error to our customers in our app.
We're able to work around this using hacks, but am curious if anyone else is seeing this, and how you've worked around it. Also, it might be good for people at docusign/akamai to look into why the performance of the www-west.docusign.net.akadns.net record is so bad.
I setup test cluster which contains 3 servers. Consul, dnsmask and NetworkManager are installed on all machines under CentOS 7.
I'd like to test simple round robin procedure:
Expected: ping consul.service.consul must send icmp requests to one of three servers.
Actual: ping always send requests to one IP address (10.82.5.6)
However ip order is changed in answer section of dig command:
[vagrant#localhost ~]$ dig consul.service.consul
; <<>> DiG 9.9.4-RedHat-9.9.4-51.el7_4.1 <<>> consul.service.consul
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 23466
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 3, AUTHORITY: 0, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
;; QUESTION SECTION:
;consul.service.consul. IN A
;; ANSWER SECTION:
consul.service.consul. 0 IN A 10.82.5.5
consul.service.consul. 0 IN A 10.82.5.4
consul.service.consul. 0 IN A 10.82.5.6
;; Query time: 2 msec
;; SERVER: 127.0.0.1#53(127.0.0.1)
;; WHEN: Wed Dec 13 13:40:20 UTC 2017
;; MSG SIZE rcvd: 98
If I reboot 10.82.5.6 node, dig returns 2 nodes and ping begins properly work - with round robin. But when I have my node 10.82.5.6 rebooted, only this node again responds to ping commands
according to https://www.consul.io/docs/agent/dns.html the DNS interface randomizes the returned nodes so it'll never be round robin.
there's also DNS caching https://www.consul.io/docs/guides/dns-cache.html the default TTL is 0, but you may have something different and/or results are cached somewhere else..
I'm building a custom DNS Server that, among other things, serves SRV records and associated A and AAAA records. I can verify that querying the server directly returns the expected answer:
$ dig lseed.bitcoinstats.com SRV #139.59.143.87 +short
10 10 9735 2c932136c294204bc65c73266300b30fe8ccb99c24fb2261d2e9980a7e8ffe9.80.lseed.bitcoinstats.com.
10 10 6331 31ce6a2b947fdbc97f10405c4062848393cf8140f33cc492aa044fe47d18f59.c6.lseed.bitcoinstats.com.
10 10 8334 283a918ae4609473c01f2e19491e9202788150dbe8d4361a3a04f3a879e9f0a.45.lseed.bitcoinstats.com.
10 10 53258 2673073e3751681b0c55aa88e5af17522c6d6b32d7d210bf4d65439d063c1ba.91.lseed.bitcoinstats.com.
However when querying through my ISPs resolver (or any of the public resolvers, like Google's 8.8.8.8) I get an empty answer back:
$ dig lseed.bitcoinstats.com SRV #8.8.8.8
; <<>> DiG 9.9.5-3ubuntu0.10-Ubuntu <<>> lseed.bitcoinstats.com SRV #8.8.8.8
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: SERVFAIL, id: 10994
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 512
;; QUESTION SECTION:
;lseed.bitcoinstats.com. IN SRV
;; Query time: 86 msec
;; SERVER: 8.8.8.8#53(8.8.8.8)
;; WHEN: Tue Nov 29 12:32:15 CET 2016
;; MSG SIZE rcvd: 51
The query returns immediately and I can see that my server receives an incoming query, however it is empty. Is it known behavior that the resolver strips SRV and additional answers? If this were the case, why is the query being forwarded at all to my server? Or is the error on my side, and the server simply replies with an incorrect answer?
Turns out that the answers were stripped due to non-matching names in the answer. The query was asking for lseed.bitcoinstats.com while the answers were replying with another domain (_lightning._tcp.lseed.bitcoinstats.com) hence the resolvers were stripping the non-matching answers, leaving just an empty reply with no answers. Setting the domain in the answers equal to the domain in the question results in resolvers passing the answers through.
When I run the following dig command on www.google.com with the +nssearch option I get no results:
mac$ dig www.google.com +nssearch
mac$
Can someone explain why no data is returned here? The +nssearch option reads the SOA of all the authoritative name servers I believe. Does this mean there are no authoritative name servers? How is that possible? The domain www.google.com obviously works so I was expecting some sort of result.
; <<>> DiG 9.9.5-3ubuntu0.2-Ubuntu <<>> www.google.com
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 40522
;; flags: qr rd ra; QUERY: 1, ANSWER: 6, AUTHORITY: 0, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
;; QUESTION SECTION:
;www.google.com. IN A
;; ANSWER SECTION:
www.google.com. 20 IN A 74.125.196.106
www.google.com. 20 IN A 74.125.196.104
www.google.com. 20 IN A 74.125.196.99
www.google.com. 20 IN A 74.125.196.147
www.google.com. 20 IN A 74.125.196.105
www.google.com. 20 IN A 74.125.196.103
;; Query time: 2 msec
;; SERVER: 192.168.186.1#53(192.168.186.1)
;; WHEN: Wed Jun 17 17:17:37 CDT 2015
;; MSG SIZE rcvd: 139
From "man dig":
+[no]nssearch
When this option is set, dig attempts to find the authoritative name servers for the zone containing the name being
looked up and display
the SOA record that each name server has for the zone.
Since there's no authority section in the response, +nssearch is going to return nothing.
www.google.com is not a zone, but a name in a zone. Therefore it doesn't have any NS records (or SOA records) for dig to display. Try dropping the www. bit and you'll get more output.