Am 26.02.20 um 18:12 schrieb Rory Campbell-Lange via Exim-users:
>
> We've never had a problem with iPhones and exim.
>
Apple did cause problems with all sorts of clients and
servers(exim,dovecot etc.. ), rangeing from desktop to mobile. In most
cases, after a while it corrects itself.
best
On Wed, Feb 26, 2020 at 05:48:09PM +0100, Luca Bertoncello via Exim-users wrote:
> > Removal of this 2-line banner seems to be a reasonable next step.
>
> As I already said, I tried to remove the 2-line banner and replace it
> with one-line-banner.
> No changes in behaviour...
OK. You also
On 26/02/20, Luca Bertoncello via Exim-users (exim-users@exim.org) wrote:
> Meanwhile it works... No idea why...
> We didn't changed something in the configuration, just deleted the profile
> in iPhone and created it again...
Perhaps you saved an outdated server certificate against the
Am 26.02.2020 um 17:31 schrieb Evgeniy Berdnikov via Exim-users:
> Handshake is definitely completed: last packet from server is
> pure Application Data, and its payload length (170) is very close to
> 146 bytes of 2-line SMTP banner (it should be slightly greater due to
> padding and hmac).
On Wed, Feb 26, 2020 at 12:27:09PM +, Jeremy Harris via Exim-users wrote:
> We're not too much closer. They agreed, during handshake, on a
> cipher-suite. We can't actually tell if the handshake completed
> (because encryption!) but the packets present make it likely.
> The client issued a
Am 26.02.2020 13:27, schrieb Jeremy Harris via Exim-users:
Hi Jeremy,
I see that's actually 8465 not 465. I assume real-465
behaves the same?
Yes, I wrote 465, but we use 8465, configured as 465...
We're not too much closer. They agreed, during handshake, on a
cipher-suite. We can't
On 25/02/2020 14:49, Luca Bertoncello via Exim-users wrote:
> Am 25.02.2020 14:57, schrieb Evgeniy Berdnikov via Exim-users:
>
>> Run traffic analyzer on the server host.
>> Post capture file here if you can't interpret output.
>
> Here the traffic dump...
I see that's actually 8465 not 465.
Am 25.02.2020 14:57, schrieb Evgeniy Berdnikov via Exim-users:
Run traffic analyzer on the server host.
Post capture file here if you can't interpret output.
Here the traffic dump...
Thanks a lot
Luca Bertoncello
(lucab...@lucabert.de)
smtp.pcap
Description: application/vnd.tcpdump.pcap
On Wed, Feb 26, 2020 at 11:11:07AM +0100, Luca Bertoncello via Exim-users wrote:
> Any other idea?
You were suggested to run traffic analyzer and post capture file to list.
--
Eugene Berdnikov
--
## List details at https://lists.exim.org/mailman/listinfo/exim-users
## Exim details at
Am 25.02.2020 16:54, schrieb Graeme Fowler via Exim-users:
Hi
A quick search (using a popular search engine) for:
exim debian stretch "error in the pull function"
...returned a significant number of results, not least of which was:
Am 25.02.2020 um 19:49 schrieb Jeremy Harris via Exim-users:
> On 25/02/2020 18:15, Luca Bertoncello via Exim-users wrote:
>> Well, but that cannot be the problem, since the iPhone can communicate
>> with Exim using the port 587 and the same configuration...
>
> In case you really mean "the same"
On 25/02/2020 18:15, Luca Bertoncello via Exim-users wrote:
> Well, but that cannot be the problem, since the iPhone can communicate
> with Exim using the port 587 and the same configuration...
In case you really mean "the same" - no, that is not expected
to work. Traditionally 587 is used for
Am 25.02.2020 um 16:54 schrieb Graeme Fowler via Exim-users:
> A quick search (using a popular search engine) for:
>
> exim debian stretch "error in the pull function"
>
> ...returned a significant number of results, not least of which was:
>
>
Am 25.02.2020 um 17:09 schrieb Heiko Schlittermann via Exim-users:
> which is fine for ESMTP, but not for SMTP. When your server issues
> the banner, it can't know if the client is able to speak/understand
> ESMTP. The server can announce it (via "EMSTP" string on the banner)
> and the client has
Am 25.02.2020 um 16:54 schrieb Graeme Fowler via Exim-users:
> A quick search (using a popular search engine) for:
>
> exim debian stretch "error in the pull function"
>
> ...returned a significant number of results, not least of which was:
>
>
Am 25.02.2020 um 17:09 schrieb Heiko Schlittermann via Exim-users:
> which is fine for ESMTP, but not for SMTP. When your server issues
> the banner, it can't know if the client is able to speak/understand
> ESMTP. The server can announce it (via "EMSTP" string on the banner)
> and the client has
Luca Bertoncello via Exim-users (Di 25 Feb 2020 16:25:03
CET):
> I have the problem with lucabert.de, too, but it is now not interesting,
> since it is my private server and no one use iPhone, here...
>
> > I wouldn't expect all mailclients understanding and handling this well.
> And I really
A quick search (using a popular search engine) for:
exim debian stretch "error in the pull function"
...returned a significant number of results, not least of which was:
https://lists.exim.org/lurker/message/20180207.150204.600285a6.en.html
The fix was to recompile with OpenSSL as the
On 25 Feb 2020, at 10:12, Heiko Schlittermann via Exim-users wrote:
If I check your mail.lucabert.de, I see a multiline response already
*before* your server has any idea, if the client would understand it.
I wouldn't expect all mailclients understanding and handling this
well.
This has
Am 25.02.2020 16:23, schrieb Jeremy Harris via Exim-users:
Mmm, that incredibly helpful error message from GnuTLS. "An error",
wow.
I thought so, too...
Best guess is that the client closed the connection before the
TLS handshake finished. Possibly it doesn't like the set of
ciphersuites
On 25/02/2020 14:47, Luca Bertoncello via Exim-users wrote:
> 15:42:24 13686 TLS: a client certificate will not be requested.
> 15:43:02 13686 LOG: MAIN
> 15:43:02 13686 TLS error on connection from [10.0.56.135]
> I=[10.0.46.11]:465 (gnutls_handshake): Error in the pull function.
Mmm, that
Am 25.02.2020 16:12, schrieb Heiko Schlittermann via Exim-users:
Hi Heiko,
Can you tell us the IP of your server? Or at least the *complete*
response you get using openssl.
Well, here is it:
depth=2 C = GB, ST = Greater Manchester, L = Salford, O = COMODO CA
Limited, CN = COMODO RSA
Luca Bertoncello via Exim-users (Di 25 Feb 2020 13:30:55
CET):
>
> I tried with openssl s_server to my server and to Google, and I see a
> difference:
>
> My server:
>
> No client certificate CA names sent
> Peer signing digest: SHA512
> Server Temp Key: ECDH, P-256, 256 bits
…
>
Can you tell
Am 25.02.2020 14:44, schrieb Jeremy Harris via Exim-users:
The main server uses Exim 4.89 from Debian Stretch paket, so I can't
update it right now...
The server with 4.92.3 is my private server (with Exim compiled). But
the problem must be solved on the server with 4.89 as primary target...
On Tue, Feb 25, 2020 at 02:14:54PM +0100, Luca Bertoncello via Exim-users wrote:
> Am 25.02.2020 14:01, schrieb Jeremy Harris via Exim-users:
> > Does the phone application provide any diagnostics?
>
> Unfortunately not...
>
> Any idea?
Run traffic analyzer on the server host.
Post capture
On 25/02/2020 13:14, Luca Bertoncello via Exim-users wrote:
> 30947 set_process_info: 30947 handling incoming connection from
> [10.0.56.135] I=[10.0.46.11]:465
> 30947 host in host_reject_connection? no (option unset)
> 30947 host in sender_unqualified_hosts? no (option unset)
> 30947 host in
Am 25.02.2020 14:01, schrieb Jeremy Harris via Exim-users:
Hi Jeremy
Session tickets are an optional thing in TLS. I doubt that
is the problem.
This was the only difference betwenn my answer and Google's one...
If you run your Exim daemon with debug enabled, what does
it show for a
On 25/02/2020 12:30, Luca Bertoncello via Exim-users wrote:
> I tried with openssl s_server to my server and to Google, and I see a
> difference:
> You see, on my server there's no "TLS session ticket"...
Session tickets are an optional thing in TLS. I doubt that
is the problem.
If you run
Hi list!
I have a server with Exim 4.89 (I tried with another server with Exim
4.92.3, too. Same problem!).
It works with all clients, but not with iPhones...
I configured it to listen on port 465 as SMTPs. If I set this port in
iPhone, it waits, and waits, and waits...
But no SMTP command
29 matches
Mail list logo
