Hi ellie,
the testsuite run was all good for the upcoming 2.4.19.
It tested folder creation, LMTP, POP3, message create/move/delete
and the ANNOTATEMORE extension for "Kolab folder types",
including access via horde / horde's IMAP library.
-> green light from me (if that counts anything).
Cheer
Hi ellie,
On Monday, 15 May 2017 08:32:30 CEST ellie timoney wrote:
> Planning to cut a release straight from the 2.4 git HEAD at time of
> writing (bfc80b3). The release tarball is already built and uploaded at
> https://cyrusimap.org/releases/cyrus-imapd-2.4.19.tar.gz
>
> If you could give tha
Hi Ellie,
On Tuesday, 09 May 2017 03:15:07 CEST ellie timoney wrote:
> I'm looking at doing a new release of 2.4 soon -- there's been a number
> of issues coming up lately which are already fixed in git, but aren't in
> a released version. It might save some support effort if these fixes
> were r
Hi Ellie,
On Tuesday, 4. October 2016 11:02:39 ellie timoney wrote:
> This version of the patch is now on master. Thanks very much for
> putting it together :)
nice :) Thanks.
> I've opened https://github.com/cyrusimap/cyrus-imapd/issues/42 for
> backporting it to 2.5 -- expecting to do it soon
Hi Ellie,
On Monday, 12. September 2016 11:35:45 ellie timoney wrote:
[clock jumps]
> Or does it? The man page says it's "not affected by discontinuous
> jumps in the system time (e.g., if the system administrator manually
> changes the clock)" -- great -- "but is affected by the incremental
> a
Hi Ellie,
On Thursday, 8. September 2016 11:07:54 ellie timoney via Cyrus-devel wrote:
> I can kind of see a point about the ability to request a monotonic
> clock, such that some sysadmin changing the system time doesn't
> prematurely timeout a bunch of idle connections. Though that doesn't
> se
On Thursday, 28. July 2016 17:16:03 Carlos Velasco via Cyrus-devel wrote:
> Looking into idled.c from v2.5.8 there is a timeout:
>
> /* Set inactivity timer (convert from minutes to seconds) */
> idle_timeout = config_getint(IMAPOPT_TIMEOUT);
> if (idle_timeout < 30) idle_timeout =
On Thursday, 28. July 2016 16:48:51 Sebastian Hagedorn wrote:
> According to the RFC, clients need to "renew" the IDLE state at least
> every 30 minutes. I would assume that the clients you see are doing that.
> If not, you might see a different issue.
I tested it a few days ago: Connected to imap
Hello,
right now cyrus imapd (2.4 / 2.5) doesn't have a timeout for connections
in the "IDLE" state.
One mail server here is serving a lot of Android K9 mail clients
that make heavy use of IMAP idle. So far around 2300 imapd
processes piled up even though it's just about 100 users in total.
Some
On Sunday, 19. June 2016 07:55:55 Bron Gondwana via Cyrus-devel wrote:
> I've also been looking at Rocks DB. Unfortunately their C wrapper (it's
> C++) doesn't seem to include the transaction support, so I'll have to
> either extend it or write my own.
>
> One major goal here, and the reason I'm
On Monday, 13. June 2016 00:05:13 Bron Gondwana via Cyrus-devel wrote:
> If twoskip is too slow (possible), then I've been quite interested in
> looking at rocksdb (http://rocksdb.org/) as an embedded engine that has
> really good performance, prefix scanning, and a good community around it.
> It'
Hi Bron,
Am 08.06.2016 um 08:22 schrieb Bron Gondwana via Cyrus-devel:
> *THE PLAN[tm]***
>
> For JMAP support, I'm going to discard the existing conversations DB and
> create a sqlite database per user which contains everything of value.
one thing to watch out for with sqlite:
It doesn't sca
Hi Bron,
given the switch to UUIDs for the folder structure in the backend of v3.0*,
I'm wondering if the following folder structure can be supported:
- Inbox/folder_below_inbox
- folder_next_to_inbox
- user/smith/other_shared_folder
Background info: Some users want to create folders below inb
Hi Ken,
Am 29.12.2015 um 16:32 schrieb Ken Murchison:
> If all that can be done safely in the signal handler is setting a global
> variable,
> I might just scrap the heartbeat functionality. The alternative is that the
> actual method processin
> code will have periodically check the status of t
On Sunday, 27. December 2015 09:48:46 Bron Gondwana via Cyrus-devel wrote:
> I think I'd like to do the next beta after CalConnect, which is Jan 10-15.
> Ken and I will both be working on making sure the calendaring is really
> solid. My main goal from this calconnect is to get our test cases up t
On Friday, 6. November 2015 18:44:52 ellie timoney via Cyrus-devel wrote:
> I think the main questions at this point are:
>
> * are we happy with the combination of --enable-objectstore to turn the
> feature on, plus --with-[backend] to select a backend, and default to
> dummy if no other backend
On Tuesday, 13. October 2015 23:03:23 Bron Gondwana wrote:
> I'm almost positive that reconstruct won't work correctly with object
> storage turned on right now.
...which reminds me that reconstruct is still broken
in the 2.4 branch since only half of my patch was applied.
The gory details:
http:
On Tuesday, 29. September 2015 08:35:34 ellie timoney wrote:
> On Mon, Sep 28, 2015, at 11:28 PM, Thomas Jarosch wrote:
> > On Monday, 28. September 2015 11:56:51 ellie timoney wrote:
> > > Hi Thomas,
> > >
> > > This is now on master. Thanks for the patch! :
On Monday, 28. September 2015 11:56:51 ellie timoney wrote:
> Hi Thomas,
>
> This is now on master. Thanks for the patch! :)
>
> ellie
thanks Ellie. Will the 2.5 backport I sent also be merged?
Cheers,
Thomas
IGTERM/select race
Also use signals_select() in prot_*() functions
Max Matveev (1):
One copy of common signal processing is enough
Thomas Jarosch (3):
Clear sigaction struct before passing it to the kernel
autoconf: Check for pselect()
idled: Add missing signals_poll() ca
This fixes broken shutdown on platforms without pselect().
A side effect of the "#ifdef HAVE_PSELECT" code path in signals_select()
is that it calls signals_poll_mask() on each run. So it will
pick up previously delivered signals.
---
imap/idled.c | 2 ++
1 file changed, 2 insertions(+)
diff --g
This fixes broken shutdown on platforms without pselect().
A side effect of the "#ifdef HAVE_PSELECT" code path in signals_select()
is that it calls signals_poll_mask() on each run. So it will
pick up previously delivered signals.
---
imap/idled.c | 2 ++
1 file changed, 2 insertions(+)
diff --g
On Thursday, 3. September 2015 22:56:08 Bron Gondwana wrote:
> I should include: the plan to fix this is to have intent logging in
> mailboxes.db - so we'll create a record for the destination name with a
> flag marking the intent to rename, such that you can't possibly create a
> rename loop.
any
On Tuesday, 1. September 2015 14:31:23 Thomas Jarosch wrote:
> --
> commit 17eb391b918c394319e4d1fe5985de10128f34d7
> Author: Greg Banks
> Date: Fri Mar 23 17:27:32 2012 +1100
>
> idle: don't use signals, use AF_UNIX dgrams
>
>
Hey there,
thanks to the recent lock debugging tool[1] and very good luck,
I was able to spot the mysterious cyrus 2.4 (and earlier) deadlock.
Here's the output from the lock debugger:
/usr/cyrus/bin/imapd (pid 3301) holding WRITE lock for
/datastore/imap-mails/user/projects/cyrus.index
/usr/c
Hi Bron,
On Friday, 7. August 2015 07:48:06 Bron Gondwana wrote:
> On Sat, Jul 25, 2015, at 01:49, Thomas Jarosch wrote:
> > Hi,
> >
> > which quota backend is the recommended backend nowadays?
> >
> > Obvious candidates are "quotalegacy" and "sk
Hi Bron,
On Sunday, 19. July 2015 10:05:12 you wrote:
> > @@ -4370,7 +4371,7 @@ static int mailbox_reconstruct_create(const char
> > *name, struct mailbox **mbptr)>
> > /* Attempt to open index */
> > r = mailbox_open_index(mailbox);
> >
> > -if (!r) r = mailbox_read_index_header(m
or "quota_db" doesn't give
any explanation which backend to prefer.
Which backend is used at Fastmail? :)
Best regards,
Thomas Jarosch
rams.
# Useful for identifying the cause of a deadlock.
# Should be executed as root for /proc access.
#
# If you identified the first hanging process,
# run "gstack pid" to get a stack trace.
#
# Licensed under the same terms as cyrus-imapd 2.4 / 2.5 / 2.6
# (c) 2015 Intra2net AG - Thomas J
Hi Bron,
the first rough version of the lock validator
is running and already has some interesting results:
The cyrus flat "db" backend doesn't do proper locking.
1. When it opens a file in myopen(), the file is not locked.
This is a common pattern in cyrus code and harmless.
myfetch() acqu
Affects 2.5 and git HEAD, patch against git HEAD.
cppcheck reported:
Dangerous usage of strncat - 3rd parameter is the maximum number of characters
to append.
---
imap/imapd.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/imap/imapd.c b/imap/imapd.c
index 687c4af..c130a59 1
Hi Bron,
On Sunday, 19. July 2015 10:05:12 you wrote:
> On Sun, Jul 19, 2015, at 07:06 AM, Thomas Jarosch wrote:
> > mailbox_release_resources() closes the file descriptor
> > to the index. This implicitly releases any lock on the file.
> >
> > Unfortunately the
mailbox_release_resources() closes the file descriptor
to the index. This implicitly releases any lock on the file.
Unfortunately the function does not reset the index_locktype state.
Therefore we didn't notice that a mailbox rename
operated on an unlocked index of the new mailbox. Yikes.
Fix it
EINTR handling in lock_unlock() was introduced
with this commit:
commit b6f2c3ad6197c4ec1da5da6a79d79c2c93ed1a1c
Author: Larry Greenfield
Date: Tue Nov 26 18:13:52 2002 +
look for EINTR on unlocks
---
lib/lock_fcntl.c | 12 ++--
lib/lock_flock.c | 4 ++--
2 files changed, 4 i
Port over Bron's fix from the fcntl() based lock_reopen() version.
Change introduced in
commit fdc0eb3d09bcc2ce916d2790c98839a61d403937
Author: Bron Gondwana
Date: Tue Jun 29 13:58:51 2010 +1000
The Rest
---
lib/lock_flock.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff
Hi Bron,
On Saturday, 16. May 2015 09:46:40 Bron Gondwana wrote:
> So the way to work through all the records in a mailbox was:
>
> int r;
> uint32_t recno;
> struct index_record record;
>
> for (recno = 1; recno <= mailbox->i.num_records; recno++) {
> r = mailbox_read_index_record(mailbox,
On Monday, 23. March 2015 23:06:16 Bron Gondwana wrote:
> > What about providing a virtual annotation that allows you to query the
> > uniqueid of a folder via IMAP? That way a client can automatically
> > include that information in debug reports if available.
>
> This already exists:
>
> C: 12
On Monday, 23. March 2015 12:09:23 Thomas Jarosch wrote:
> A tool that lists the user folder -> uniqueid based path
> in a machine parsable way (=scriptable) might help here.
after talking to a colleague in development, another idea came up:
What about providing a virtual annotation th
Hi Bron,
On Saturday, 21. March 2015 06:32:39 Bron Gondwana wrote:
> Pretty sure we've already had this dance. I did build a half-way-house
> solution a few years ago, which looked like this:
>
> b/user/brong/user.brong.subfolder/
>
> It had the advantage that no email folder contained another
Hi,
we are doing backups of our imap mailboxes using
simple unix tools like find + afio. Sometimes, when a mailbox
is still locked, deleted files were not expunged immediately
and therefore ended up in the backup.
During testing we found an additional problem with re-creating
a deleted folder and
Hi Bron,
On Tuesday, 17. March 2015 09:50:34 Bron Gondwana wrote:
> Bron:
> - working on uniqueid-based-paths
> - we'll need upgrade path for existing installs (maybe detect old path
> and rename-on-open?) - reconstruct et al will need support for the new
> paths. Rebuild mailboxes.db from the
On Friday, 12. December 2014 11:07:39 you wrote:
> Is not possible to user strmove instead of strcpy?
POSIX doesn't support strmove() AFAIK.
-> it's non-standard.
Thomas
Hi Bron,
I just noticed an important bugfix for ACL removal is still
sitting in the bug tracker untouched.
Please have a look at
https://bugzilla.cyrusimap.org/show_bug.cgi?id=3824
Without it the ACL strings get messed up in mailboxes.db
on current glibc versions. We run the patch for about a ye
Hi,
I just discovered an interesting behavior and I'm wondering
why cyrus-imapd 2.4.17 behaves the way it does:
Basically I do this:
- Select a mailbox
- Get the list of UIDs via "UID SEARCH 1:*"
- Add a new message in another client / connection
- Get the list of UIDs again via "UID SEARCH 1:*"
On Thursday, 27. February 2014 09:42:57 Thomas Jarosch wrote:
> Hi,
>
> we run the 'quota' command every fifteen minutes
> to gather statistics about the disk usage.
>
> After updating to cyrus-imapd 2.4.x, I sometimes (rarely though)
> get this message in the
Hi,
we run the 'quota' command every fifteen minutes
to gather statistics about the disk usage.
After updating to cyrus-imapd 2.4.x, I sometimes (rarely though)
get this message in the logs:
Feb 27 01:30:02 mail quota[12433]: failed building quota list for '': System
I/O error: No such file or
**
Was it intentional that the CLOSED response should always be sent?
If yes, what was the intention?
May be the wording can be clarified a bit before the updated RFC is final:
Should the server send the CLOSED response even if QRESYNC is not enabled?
Best regards,
Thomas Jarosch
On Friday, 7. February 2014 08:04:55 Bron Gondwana wrote:
> > @Bron: Do you know by chance if this error case is detected in 2.5?
> > I took a quick look at the code in git HEAD and it seems to behave the
> > same.
> I was pretty sure that it detected that case - but I may be wrong. I'll
> take a
Hi,
I just encountered one system where cyr_expire / reconstruct / imapd process
died during the index upgrade from cyrus 2.3.16 to 2.4.17.
The error message was:
"fatal error: failed to mmap new message file"
To make a long debug session short: One folder had a broken "cyrus.index"
file th
Hi Bron,
On Thursday, 2. January 2014 09:52:13 Bron Gondwana wrote:
> user.foo.sub %(NAMELOCK user.foo.sub TYPE RENAMETEMP)
I just came up with a case that fails at least in cyrus 2.4:
Rename a user while a mailbox is currently in SELECT state.
Cyrus 2.4 will execute the mailbox rename,
Hi Bron,
On Thursday, 2. January 2014 09:52:13 Bron Gondwana wrote:
> At this point, we go through all the old folder and delete the on-disk
> files. Once that's done, we can do a single update in the mailboxes.db:
>
> user.foo.sub %(TYPE DELETED)
>
> (we keep DELETED tombstones in mail
Hi Bron,
I'm currently testing the upgrade path from 2.3.16 to 2.4.17.
Everything is smooth so far, mailbox indexes were upgraded
automatically after starting 2.4.17 for the first time:
--
Dec 18 13:48:00 ctl_cyrusdb[14851]: checkpointing cyrus databases
Dec 18 13:48:00 ct
On Wednesday, 1. May 2013 00:03:20 Bron Gondwana wrote:
> What are your main blockers to moving to 2.4? I suspect the reindexing IO
> his is a major factor, and something I'd like to solve (by delaying the
> inevitable really... there are enough cache format changes that
> reindexing will be neces
Hi Bron,
On Tuesday, 30. April 2013 08:53:24 Bron Gondwana wrote:
> Hi - have you looked at all at the special-use support in mainline Cyrus?
>
> The xlist-* behaviour is planned to be removed, in favour of using the RFC
> 6154 mandated annotation.
>
> The implementation in git at the moment doe
On Thursday, 13. October 2011 15:48:38 Jeroen van Meeuwen (Kolab Systems)
wrote:
> I was able to successfully create a new ticket:
>
> https://bugzilla.cyrusimap.org/show_bug.cgi?id=3580
Thanks for testing!
Thomas
On Thursday, 13. October 2011 12:15:17 Dave McMurtrie wrote:
> > @Dave: Any issue known with bugzilla?
>
> Not that I'm aware of. Why do you ask? If there's something broken, let
> me know and I'll take care of it.
I tried to create a new bug (twice) and it started processing after hitting
"su
On Thursday, 13. October 2011 11:18:33 Bron Gondwana wrote:
> > What about the actual code change?
>
> The code looks good, thanks. Do you want me to push it to git master?
Yes, pretty please.
If the two tab changes are an isuse, I can send a fixed up patch.
@Dave: Any issue known with bugzill
Hi Bron,
On Thursday, 13. October 2011 11:03:16 you wrote:
> On 10/13/2011 10:13 AM, Thomas Jarosch wrote:
> > Also don't mix tabs and spaces.
>
> Please don't do this. I know the tabs/spaces standard is ugly, but
> that's what this codebase has.
Normally I
On Thursday, 13. October 2011 10:13:54 Thomas Jarosch wrote:
> Also don't mix tabs and spaces.
I guess my mail client will eat the tabs in the patch.
Here's the patch as file for easy "git am" consumption.
Cheers,
Thomas
From 6251ba2071e6adf05f630608cbf0768417ad9d3f Mon Se
Also don't mix tabs and spaces.
Creation of a new bug on cyrus bugzilla currently
doesn't work for me (blank page on submit).
Signed-off-by: Thomas Jarosch
---
timsieved/actions.c | 17 +++--
1 files changed, 11 insertions(+), 6 deletions(-)
diff --git a/timsieved/a
ts with the patch back then.
Anyhow we're going to fix this.
@Bron: Do you have any vague plans in mind
for a 2.4.11 release date?
Best regards,
Thomas Jarosch
Hi Дилян,
here's some feedback about your build system question.
Note: I'm not one of the cyrus core developers.
> if I rewrite the build system of Cyrus imap 2.4(.10) to use Automake to
> generate the Makefile.in-files, will the patch be accepted in reasonable
> time in git/master?
Have you con
of Technology and Economics, Hungary.
Best regards,
Thomas Jarosch
On Thursday, 14. April 2011 17:45:11 Thomas Jarosch wrote:
> Here's my proposal to make duplicate suppression more robust:
> Combine the "message-id" header field with the "date" header field
> for the duplicate check. If a broken client reuses the message-id
;s my proposal to make duplicate suppression more robust:
Combine the "message-id" header field with the "date" header field
for the duplicate check. If a broken client reuses the message-id,
there's a good chance the date field changed.
Opinions?
Best regards,
Thomas Jarosch
Hello Bron,
On Tuesday, 17. November 2009 13:28:49 Bron Gondwana wrote:
> Does anybody out there use annotations much? Does anybody know any code
> that would be broken by changing the way annotations are done?
The Kolab project and the horde framework counterparts use it to store
the type of a
On Wednesday, 9. September 2009 18:43:43 Dave McMurtrie wrote:
> > TJ> Regarding the buffer overflow: The cert website currently outputs a
> > TJ> "Lotus Notes exception". Is the overflow theoretically exploitable
> > TJ> via a malicious email or does a user need to upload a malicious
> > TJ> sieve
On Wednesday, 9. September 2009 15:47:14 Ken Murchison wrote:
> I'd like to announce the releases of Cyrus IMAPd 2.2.13p1 and 2.3.15.
> These releases should both be considered production quality. These
> releases are being made at this time to fix the potential buffer
> overflow vulnerability des
On Tuesday, 24. March 2009 18:21:39 Wesley Craig wrote:
> On 24 Mar 2009, at 06:47, Thomas Jarosch wrote:
> > Thanks for calling. I don't like to spoil the party,
> > what about the issue which allows to create empty ACLs?
> >
> > I seem to have forgotten to op
Hi Ken,
On Tuesday, 24. March 2009 11:32:10 Ken Murchison wrote:
> If there are any outstanding issues that you believe still need to be
> addressed in 2.3.14, please let me know ASAP.
Thanks for calling. I don't like to spoil the party,
what about the issue which allows to create empty ACLs?
Th
Hi Wes,
rfc4314 seems to specifically disallow empty identifiers. Also, I think
you patch would probably permit an identifier of "-".
The check is done after the "-" handling so it should take care of it.
BTW, I have a patch to this code that I'm currently holding, which introduces a
lead
Hello together,
attached is a small patch for discussion. It prevents "setacl"
for empty indentifiers.
If I read RFC 2086 correctly, empty identifiers seem to be allowed
(an oversight?), but most clients won't be able to handle this ACL
and there is also the question if there is a valid use case
Bron,
I've quickly reviewed your patches and put them in our production server.
Seems to work fine :-)
Maybe the comment of the "cyrus-always-unlink-2.3.12.diff" patch can be
expanded a bit so I won't be removed in the future again: unlink() is normally
not needed as fopen(stagefile, "w+") will
Hi Bron,
On Thursday, 11. September 2008 03:33:15 you wrote:
> Can you post the logic that your script was using? That might help.
The problem is triggered by a perl script that syncs a list of users
against the actual user list in cyrus. It uses perl-IMAP-Admin to do this.
Unknown users get de
On Thursday, 11. September 2008 14:40:03 you wrote:
> We are using Linux and
>
> uname -a
> Linux ms01 2.6.5-7.276.PTF.205323.0-bigsmp #1 SMP Mon Jul 24 10:45:31
> UTC 2006 i686
>
> At http://www.intevation.de/roundup/kolab/issue840 we saw that someone
> recomended
> a linux 2.6.19.2 http://www.in
Hello Henryk,
which version of cyrus-imapd are you using?
Which operation system are you using?
If it's Linux, what kernel version do you use?
In the past there was a bug in the kernel (mmap)
that could cause skiplist corruption.
Thomas
Hello Bron and Wes,
I've experienced another bug and I'm not sure if the skiplist/locking code is
at fault or not. I was deleting huge amounts of mail in many subfolders
on a slow box. The tool that started the request got a timeout after some
minutes and this left cyrus-imapd in a broken state:
Hello together,
currently unix sockets get created by cyrus-master
with ownership of "root.root" and file mode 0777.
Attached patch makes the user, group and file mode configurable.
If nothing is specified in cyrus.conf, it defaults to CYRUS_USER
(+group of the user) and mode 660 for improved sec
Hello Ken,
attached is a small patch that sets the mtime of the corresponding file to the
INTERNALDATE of a mail. We use a perl script to restore single mailboxes
from backup via IMAP and this enables us to restore the arrival time, too.
This was the last one from my patch queue ;-)
Thomas
--
uot;
and delete all your mail, so I changed it to "-M mailbox".
The verbose output was also improved to show which folders were skipped.
The patch was developed for cyrus-imapd 2.2.12 and applies cleanly to HEAD.
Is this mailinglist or bugzilla t
Hello Ken,
attached is a small patch to make "got_signal" volatile. My local SVN history
says it's a patch from Debian created in 2003. This should make signal
handling more robust in case the value of got_signal gets cached
in a register. Applies fine to cyrus-imapd 2.3.12p2.
Thomas
--- cyrus
Hello Ken,
I've noticed a little piece of code and wanted to ask about the original idea
behind it. In mycanconifyid() is a special code path if the identifier begins
with "group:". If so, we call getgrnam() and then copy the resulting group
name into the buffer.
I'm wondering why the code doe
82 matches
Mail list logo
