Closed lmeunier closed 7 months ago
@mjl- in the end, I reused the retry mechanism already present in the code for errors received at the start of the connexion. No changes in the configuration are needed. I don't know if it's better, or not, compared to https://github.com/mjl-/mox/issues/149#issuecomment-2041007381
I'm also wondering how to write a test for this commit. Any idea will be appreciated.
this change also seems useful, or did you run into problems with this?
I didn't encounter any problems during my tests with this PR. But, I would prefer to use the other PR for these reasons:
RCPT
command is rejected, then mox try agin a few minutes later with ipv4, and finaly the mail is deliveredthis change didn't look too dangerous to me wrt changing existing behaviour. and it will be useful to also detect this error code during smtp rcpt to in case of dns block listings of ipv6 vs ipv4, so i merged it, thanks!
mox was already giving another try for received errors after the
HELO
/EHLO
command. Now mox do the same for received errors when trying to deliver the message to the remote SMTP server.This should help to deliver messages to SMTP server that rejects incoming messages because of bad ipv4 or ipv6 configuration (for example for servers checking reverse DNS records). mox will now try to deliver messages on both ip family instead before considering the error as permanent.
fix #149