Bad port for federation, bug good configuration

On v3.18.1 in docker, I configured DNS SRV and TXT records for federation.
So now i'm pretty OK, but for unknown reasons, rocketchat interface explains to me that the port is misconfigured:

The test fails:

In server log I just got every minutes:

server.js:204 Federation ➔ http.error Error Error: ETIMEDOUT

But the thing is, from the host server (and same thing from inside the docker container), DNS configuration seems to be OK:

# dig srv _rocketchat._https.chat.axiom-team.fr

; <<>> DiG 9.10.3-P4-Debian <<>> srv _rocketchat._https.chat.axiom-team.fr
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 25927
;; flags: qr rd ra ad; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 512
;; QUESTION SECTION:
;_rocketchat._https.chat.axiom-team.fr. IN SRV

;; ANSWER SECTION:
_rocketchat._https.chat.axiom-team.fr. 60 IN SRV 1 1 443 chat.axiom-team.fr.

;; Query time: 21 msec
;; SERVER: 8.8.8.8#53(8.8.8.8)
;; WHEN: Tue Sep 28 19:02:10 CEST 2021
;; MSG SIZE  rcvd: 104

We can see the port here is 443 as well.
Any Ideas ? How can I dig the problem ?

Start at the federation error reported at server.js:204 and reproduce the timeout using the provided DNS SRV/TXT configuration inside the Docker container. Compare the configured port with the federation connection behavior and document or fix the cause so the federation test succeeds without recurring ETIMEDOUT errors.