1547358 - Turn off stateless DNS

Status: NEW

(Taskcluster :: General, task)

Description

[Dustin J. Mitchell [:dustin] (he/him)]

Once everything that ever used livelogs is using websocktunnel, we can turn off the stateless DNS service and park the taskcluster-worker.net domain.

Comment 1

[Dustin J. Mitchell [:dustin] (he/him)]

This is pending the generic-worker -> docker-worker transition. I don't see a bug to track that at the moment, though -- Pete, does one exist?

Comment 2

[Pete Moore [:pmoore][:pete]]

Indeed, but 1499055 is the correct bug.

Comment 3

[Dustin J. Mitchell [:dustin] (he/him)]

Docker-worker is still using stateless DNS, and we're still using docker-worker :(

Comment 4

[Chris Cooper [:coop] (he/him)]

Dustin: with docker-worker persisting for some TBD amount of time, where do we stand with stateless DNS now?

Do we need to redeploy it? Should we replace it with websocktunnel in docker-worker?

Comment 5

[Dustin J. Mitchell [:dustin] (he/him)]

Do we need to redeploy it? Should we replace it with websocktunnel in docker-worker?

Maybe -- that's bug 1566940.

Comment 6

[Dustin J. Mitchell [:dustin] (he/him)]

As long as docker-worker lives, this needs to live. Which seems OK. Also, bump in triage :)

Comment 7

[Chris Cooper [:coop] (he/him)]

As an offshoot of bug 1602932, I looked at the papertrail logs for stateless DNS. We have a second stateless DNS server configured in eu-west-1 that never gets used AFAICT, or at least it never logs anything to papertrail.

Note that this is the only instance we have running in eu-west-1. Perhaps this server needs to exist in eu-central-1 where we do actually have worker instances running to be useful? I don't know enough about how these servers are setup to be able to answer that.

Comment 8

[Dustin J. Mitchell [:dustin] (he/him)]

dustin@rubin ~ $ dig +norecurse -tns taskcluster-worker.net @2001:503:231d::2:30

; <<>> DiG 9.11.14-RedHat-9.11.14-2.fc31 <<>> +norecurse -tns taskcluster-worker.net @2001:503:231d::2:30
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 43410
;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 2, ADDITIONAL: 3

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
;; QUESTION SECTION:
;taskcluster-worker.net.                IN      NS

;; AUTHORITY SECTION:
taskcluster-worker.net. 172800  IN      NS      ns1.taskcluster-worker.net.
taskcluster-worker.net. 172800  IN      NS      ns2.taskcluster-worker.net.

;; ADDITIONAL SECTION:
ns1.taskcluster-worker.net. 172800 IN   A       52.49.228.53
ns2.taskcluster-worker.net. 172800 IN   A       34.213.136.122

It doesn't matter where the workers are -- it's where the users are that matters. I suspect that one of those ns's above is the eu-west-1 instance?

Comment 9

[Chris Cooper [:coop] (he/him)]

(In reply to Dustin J. Mitchell [:dustin] (he/him) from comment #8)

It doesn't matter where the workers are -- it's where the users are that matters. I suspect that one of those ns's above is the eu-west-1 instance?

Yes, 34.213.136.122 is in eu-west-1 and 52.49.228.53 is in us-west-2.

Comment 10

[Dustin J. Mitchell [:dustin] (he/him)]

Interesting! So maybe it's not logging because it's broken, or because its logging config is broken.

Comment 11

[Pete Moore [:pmoore][:pete]]

Note, comment 6 still holds - we can't close this bug until we've migrated off of docker-worker.