1 |
On 01/19/2018 12:48 PM, Grant Edwards wrote: |
2 |
> Yep, and it looks like the Postfix equivalent is a custom pipe transport. |
3 |
> Once you know what phrases to google for, it's a lot easier. |
4 |
|
5 |
*nod* |
6 |
|
7 |
I figured that you would be able to find something. |
8 |
|
9 |
Hence why I mentioned the terms. ;-) |
10 |
|
11 |
> I could live with queueing/retrying as long as the eventual failures |
12 |
> generated messages that are sent back to the sender. Those failure |
13 |
> messages would need to be sent via a normal SMTP smarthost/relayhost |
14 |
> (with AUTH) and not via the custom mailer. |
15 |
|
16 |
I would expect that it is possible to fulfill those requirements. |
17 |
|
18 |
> Yes, I've been thinking about that. I think I'll try that first -- |
19 |
> if my understanding of the failure mode is correct, it should work. |
20 |
|
21 |
The simpler solution is usually nicer. |
22 |
|
23 |
> I wrote the server I'm using now, but it uses somebody else's snmpd |
24 |
> module, and that's where the SSL breakage is. I've filed a bug, and I've |
25 |
> been doing some reading toward attempting a fix, but it looks like it |
26 |
> might be a bit hairy: it involves Python's asyncore/asynchat framework |
27 |
> (and process pools). What's missing is handling for ssl "want read" |
28 |
> and "want write" exceptions. |
29 |
|
30 |
"snmpd" or "smtpd"? |
31 |
|
32 |
You lost me at Python. (I know it's a personal prejudice. But I think |
33 |
I'm allowed to have it as long as I acknowledge them as such.) |
34 |
|
35 |
|
36 |
|
37 |
-- |
38 |
Grant. . . . |
39 |
unix || die |