2

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

CHEM_Eugene
  • 438
  • 3
  • 20

1 Answers1

2

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..

Gal Ben-Haim
  • 17,433
  • 22
  • 78
  • 131