itory still points at git.aox.org.
>
> Goodbyte, Gerd.
> --
> The last thing one knows in constructing a work is what to put first.
> -- Blaise Pascal
>
>
--
Mark Felder
f...@feld.me
On Sun, Jan 19, 2020, at 08:47, Mark Felder wrote:
> Seeing this with info logs which I never had on in the past:
>
> Archiveopteryx[32340]: 61158/13/1 Note: Expected type string for column
> "flag", but received null
>
Actually I think it's the mailboxes table whic
l_mu" btree (mailbox, uid)
Foreign-key constraints:
"flags_flag_fkey" FOREIGN KEY (flag) REFERENCES flag_names(id)
"flags_mailbox_fkey" FOREIGN KEY (mailbox, uid) REFERENCES
mailbox_messages(mailbox, uid) ON DELETE CASCADE
--
Mark Felder
f...@feld.me
On Wed, Dec 11, 2019, at 21:20, Tom Ivar Helbekkmo wrote:
> Mark Felder writes:
>
> > I suppose this is rarely-tested functionality but it looks like aox
> > cannot connect to a postgres unix socket.
>
> That's because it chroots into the "jail" direct
default value (64) is not provided in the config file. I suppose I would have
never composed my original email if that was included in the man page with a
description of which internal behaviors it affects.
--
Mark Felder
f...@feld.me
ther gmail’s odd size limit comes from a similar
> calculation. It’s 37.x megabytes IIRC, don’t remember the x).
>
> Arnt
>
>
Did this ever get merged? I can't seem to find any commits that match this
description. I looked back through 2016, before this message was composed.
--
Mark Felder
f...@feld.me
, c, d;
^
core/md5.cpp:247:5: error: 'register' storage class specifier is deprecated and
incompatible with C++17 [-Werror,-Wdeprecated-register]
register uint32 a, b, c, d;
^
5 errors generated.
--
Mark Felder
f...@feld.me
osure" (\Seen $X-ME-Annot-2 $IsMailingList
$HasAttachment) "19-Oct-2016 17:13:55 -0400" {16325504}': 764851 BAD
Literal too big: 16325504>960
What is causing this? imapsync, postgres, or aox?
--
Mark Felder
f...@feld.me
; <verylon...@mail.gmail.com>>*.
>
> Thanks. I'm looking forward to playing with this more.
>
>
Did that email come from GMail and have multiple labels on it? If so, to
aox it might look like multiple emails stored in different folders. It
should be deduplicated within aox, but I could understand why it would
output multiple copies with aox export.
--
Mark Felder
f...@feld.me
[-Werror,-Wconstant-conversion]
| SSL_OP_NO_SSLv2
^
1 error generated.
--
Mark Felder
f...@feld.me
ps I will just make the postgresql-server always
optional.
--
Mark Felder
f...@feld.me
n a
"release".
https://github.com/aox/aox/commit/4e73c54509612104f0afb9d968b5e851998d14e3
--
Mark Felder
f...@feld.me
ize which a process may lock into
memory
using mlock(2)
Now I need to find out why 64K is a sane default.
--
Mark Felder
f...@feld.me
On Tue, Dec 23, 2014, at 08:34, Arnt Gulbrandsen wrote:
I should elaborate...
IMAP makes the server do a number of things that don't really serve the
user, and that most clients don't use or use very little, but that cannot
be skipped because a significant number of clients rely on them
Would it be possible to put the aox.org website files on github or the like?
It's very much out of date. Tons of information is specific told older
versions. Still mentions features such as views which were removed in 3.2.0,
references to Postgres 8.x which isn't compatible anymore, etc. If the
Happened to correlate this in my mail logs to the event at hand:
Jul 25 21:09:42 staffmail Archiveopteryx: 19547/6 die( Invariant ) called
Jul 25 21:09:42 staffmail Archiveopteryx: 19547/6 Invariant failed while
processing Query::notify()
testing outbound again
July 15 2014 1:53 PM, Mark Felder wrote:
testing outbound again
I have no idea how this address got selected. Sorry for the noise...
On May 29, 2014, at 12:59, Axel Rau axel@chaos1.de wrote:
Multi-Master-Backend won’t work with aox, says Arnt. Has to do with the
complexity of the IMAP protocol.
Postgres-XC with fast interconnect is supposed to be near instant replication.
I figured properly using database
On 2014-05-03 16:00, ja...@mansionfamily.plus.com wrote:
On 29/04/2014 13:09, Mark Felder wrote:
It has been fantastic; I can't recommend it enough.
My experience upgrading from 3.1.8 today was not so good.
I am running FreeBSD 9.2, Postgres 9.2 from ports.
Hi James,
I'm the FreeBSD
On 2014-05-05 07:24, Arnt Gulbrandsen wrote:
On Saturday, May 3, 2014 11:00:17 PM CEST, ja...@mansionfamily.plus.com
wrote:
My Postgres 9.2.8 pg_stat_activity definitely uses 'pid', not
'procpid'.
I've changed that to 90200 now. Unless Abhijit disagrees?
This change did happen in 9.2. I
On 2014-04-29 02:58, Stephen R. van den Berg wrote:
Since the release of 3.2.0 I have not had aox crash
or memory leak on me even once, which is exceptional.
Congrats to all contributors.
It has been fantastic; I can't recommend it enough.
GMail has an IMAP feature for clients that never seem to want to expunge
called auto-expunge. I've been using Apple mail clients (Mail.app, iOS
Mail) a lot lately and I find it annoying that when I delete messages
they are only marked as deleted and not expunged. When I log into my
webmail
Doesn't seem to be playing nice as a clean install on freebsd with
postgres 9.3.
Starting clean, no database or users:
root@aox:/var/log # /usr/local/libexec/aox/installer -p pgsql -a 1.2.3.4
Connecting to Postgres server 1.2.3.4 as user pgsql.
Creating the 'aox' PostgreSQL user.
Creating the
On Mon, Feb 17, 2014, at 9:40, Stephen R. van den Berg wrote:
Mark Felder wrote:
Actually a git tag would be sufficient as well; I can pull directly from
Github with FreeBSD's ports system
Simply get the tip of the master branch.
It's the best version so far.
I'm the maintainer of mail
On Fri, Feb 14, 2014, at 7:18, Mark Felder wrote:
On Wed, Jan 15, 2014, at 8:12, Arnt Gulbrandsen wrote:
Abhijit will make the tarball, which may take him a few days or weeks.
Still waiting on that tarball :-)
Actually a git tag would be sufficient as well; I can pull directly
On Wed, Jan 15, 2014, at 8:12, Arnt Gulbrandsen wrote:
Abhijit will make the tarball, which may take him a few days or weeks.
Still waiting on that tarball :-)
On Thu, Oct 10, 2013, at 4:15, Abhijit Menon-Sen wrote:
The list was down briefly. Is it back up now?
-- ams
most certainly is :)
I feel like I'm crying wolf!, so let's put this bug on ice for a
moment: my previous hang issues were a showstopper for me and I got
really busy at work. I migrated my email out of aox because I didn't
have time to debug further.
Last night I saw that a release was likely approaching and I'm the
Will THREAD be enabled before 3.1.4, or is it still considered too
experimental?
// enable these when the tb problem is resolved
//c.append( THREAD=ORDEREDSUBJECT );
//c.append( THREAD=REFS );
//c.append( THREAD=REFERENCES );
On Thu, Aug 29, 2013, at 9:58, Axel Rau wrote:
Am 29.08.2013 um 03:34 schrieb Mark Felder f...@feld.me:
git head is still broken on FreeBSD. Connections just hang/stall forever
without any obvious cause. My testing has been on FreeBSD 9.x
specifically, but I'll throw together a 10
On Thu, Aug 29, 2013, at 14:58, Arnt Gulbrandsen wrote:
Axel Rau axel@chaos1.de writes:
In my nginx.conf, I have
ssl_ciphers HIGH:!aNULL:!MD5;
which produces PFS.
I added it now, but did not commit.
current recommended cipher settings that defeat BEAST attack look like
this:
On Thu, 17 Jan 2013 08:17:19 -0600, Axel Rau axel@chaos1.de wrote:
That's pretty fast.
How many spindles are in use by the db?
I'm getting with fbsd 8.2, pg 9.2.2 on ufs2:
Hilariously this server has a ZFS mirror with one 250GB SATA drive and one
250GB IDE drive. I have 750GB SATA
Running commit 395c922825addc713e4b53306e93cf5f259ba1ef and my connections
aren't dying. I know you've implemented a dirty workaround, but thanks!
On 31.12.2012 12:04, Arnt Gulbrandsen wrote:
Are the crashes related to memory overuse? If yes, I am testing a fix
right now, if no, I heard one can get 9.1 AWS nodes easily ;)
I'm really not sure what the cause is. I haven't put aox into debugging
to look further. I'd be willing to test
Hi all,
Tried running the latest git and messages were getting stuck in my exim queue.
A restart of aox would allow a flush to complete. Has anyone else seen this?
Here's a weird idea but might work:
Modify your webmail program to try to fetch the attachment from the filesystem.
If this fails, fall back to pulling it straight from the database. You then
have a task (a trigger? a cron job?) that exports the big attachments from the
database to the
I can't comment about scaling to this size, but it should be possible. However,
I do wonder why you want to move attachments out of the database? You do
realize that the attachments are also deduplicated, right? Send a 1MB
attachment to a million users on your service and it will only be stored
On Tue, 04 Sep 2012 13:23:39 -0500, Arnt Gulbrandsen
a...@gulbrandsen.priv.no wrote:
Show me what you would like to see?
Arnt
Dovecot's logs are nice: (I've sanitized the users)
mailbox# tail -f maillog | grep -i imap
Sep 6 06:54:51 mailbox dovecot: auth(default): client in: AUTH
On Fri, 31 Aug 2012 13:56:08 -0500, Arnt Gulbrandsen
a...@gulbrandsen.priv.no wrote:
There is a table you might like, probably called connections.
Yes, however the fail2ban utility (and others like it) only read from log
files. I suppose I could write my own script to scrape the data out
Currently it seems aox doesn't do any useful logging of user connections.
I'd really appreciate if IMAP/POP/SMTP Submit connections were logged with
the user's IP as well as whether it was successful or not. When I enable
INFO logging this is as verbose as I'm getting:
Aug 31 09:00:48 mwi1
On Tue, 14 Aug 2012 05:14:08 -0500, Thufir hawat.thu...@gmail.com wrote:
I was looking at your github page and see that you're still active, but
3.1.3 was released on 2010-03-10. You're still active?
There's definitely activity in the git repo. :-)
Thanks for the patch. I've made a modified version for the ports tree. I
was unsuccessful just telling the deliver/Jamfile to do aoxdeliver :
deliver.cpp. It would error for some reason at the end of the build so I
just went ahead and did the whole aoxdeliver.cpp creation.
Patch can be
Thanks, was just checking. I hadn't taken the time to investigate its real
purpose.
Keep up the good work. This is probably my favorite open source project :-)
On Tue, 05 Jun 2012 05:47:35 -0500, Arnt Gulbrandsen
a...@gulbrandsen.priv.no wrote:
The conflict is in a sense appropriate. Our deliver does exactly the
same thing as e.g. Cyrus, so you should have only one.
I suggest renaming it as aoxdeliver. Will do it upstream (I have a
collection
Hi all,
I'm currently the maintainer of Archiveopteryx on FreeBSD and I've
received notification that the deliver binary conflicts with Postfix and
another MTA named Avenger (mailavenger.org). This means that
installing Archiveoptertx as a storage backend for either of these MTAs
would
On Mon, 26 Mar 2012 05:47:56 -0500, tmr tm...@o2.pl wrote:
Setting this and after logging I can still remove messages.
Correct, it's not read-only. The effect should be that even if a
deletion is attempted it doesn't really work because a refresh of the
mailbox should show the message
On Sat, 10 Mar 2012 12:34:19 -0600, ja...@mansionfamily.plus.com wrote:
What's the golden source at the moment? Is the current release still
3.1.3 (2010-03-10)?
James
Git checkouts work fine. There are a few more things on the TODO list for
3.1.4 and they're no small tasks which is
On 01.11.2011 21:32, f...@feld.me wrote:
I've been running it that way since you offered up the fix without
any side
effects anyway...
Oh dear...
I humbly apologize, but that was like the first email I sent from my
touchpad. I've now learned that it sends HTML emails by default and
I
On Wed, 10 Aug 2011 04:46:52 -0500, Arnt Gulbrandsen
a...@gulbrandsen.priv.no wrote:
Odd warning, even 65535 isn't really a large number, but the code is
broken. Change ntohs(1) to just 1.
Just want to mention that this DOES fix the problem, so go ahead and throw
it in git
unless it's
Running a git checkout from last night and today I saw the cpu spike. It's
just chewing up 70-90% cpu and doing nothing.
truss output gave me this over and over and over:
85452: gettimeofday({1312920645.853700 },0x0)= 0 (0x0)
85452: clock_gettime(13,{1312920645.0 }) = 0 (0x0)
51 matches
Mail list logo