[Dovecot] v2.1.rc5 (85a9b5236b6c) Error: lmtp client: DNS lookup of $FQDN failed: connect(dns-client) failed: No such file or directory
When the Sieve plugin tries to send a vacation message or redirect a message to another address it fails. dovecot: lmtp(6412, u...@example.com): Error: lmtp client: DNS lookup of orange.example.com failed: connect(dns-client) failed: No such file or directory dovecot: lmtp(6412, u...@example.com): Error: dAIOClYSJE8MGQAAhQ0vrQ: sieve: msgid=4f241255.2060...@example.com: failed to redirect message to admin+u...@example.com (refer to server log for more information) But the dns-client sockets are created when Dovecot starts up: # find /usr/local/var/run/dovecot -name dns-client -exec ls -l {} + srw-rw-rw- 1 root staff 0 Jan 28 16:15 /usr/local/var/run/dovecot/dns-client srw-rw-rw- 1 root root 0 Jan 28 16:15 /usr/local/var/run/dovecot/login/dns-client Hum, is it Dovecot or Pigeonhole (dovecot-2.1-pigeonhole 1600:b2a456e15ed5)? Regards, Pascal -- The trapper recommends today: c01dcofe.1202...@localdomain.org
[Dovecot] maildir vs mdbox
Hi, I am planning on running on test between maildir and mdbox to see which is a better fit for my use case. And I'm just looking for general advice/recommendation. I will post any results I obtain here. Important question: I have multiple users hitting the same email account at the same time. Can be a problem with mdbox? (either via thunderbird or with custom webmail apps). I remember having huge issues with mbox a decade ago because of this. Maildir fixed this... will mdbox reintroduce this problem? This is a very important point for me. Here is my use case: - Ubuntu server (any specific recommandations on FS to use?) - Standard PC hardware (core i5 or i7, few gigs of ram, hdds at first, probably ssd afterwards, nothing very fancy) - Serving only a hand full of email accounts but some of the accounst have over 3 millions emails in them (with individual mail folders having 100k+ emails) - Will use latest dovecot (2.1 when it comes out) - fts-lucene or fts-solr? -jd
Re: [Dovecot] maildir vs mdbox
On 28.1.2012, at 17.59, Jean-Daniel Beaubien wrote: I am planning on running on test between maildir and mdbox to see which is a better fit for my use case. And I'm just looking for general advice/recommendation. I will post any results I obtain here. Maildir is good for reliability, since it's just about impossible to corrupt, and even in case of filesystem corruption it's easier to recover than other formats. mdbox is good if you want the best performance. Important question: I have multiple users hitting the same email account at the same time. Can be a problem with mdbox? No problem. - Serving only a hand full of email accounts but some of the accounst have over 3 millions emails in them (with individual mail folders having 100k+ emails) Maildir gets slow with that many mails in one folder. - fts-lucene or fts-solr? fts-lucene uses the latest CLucene version, which is a little old. With fts-solr you can use the latest Solr/Lucene. So as long as you don't mind setting up a Solr instance it should be better. The good thing about fts-lucene is that you can simply enable it and it works without any external servers.
Re: [Dovecot] maildir vs mdbox
Wow, incredible response time :) I have 1 more question which I forgot to put in the initial post. Considering my use case (small number of accounts but alot of emails per account, and I should add that they are mostly small emails, most under 5k, alot under 30k) what mdbox setting would you recommend i start testing with (mdbox_rotate_size and mdbox_rotate_interval). -JD On Sat, Jan 28, 2012 at 11:05 AM, Timo Sirainen t...@iki.fi wrote: On 28.1.2012, at 17.59, Jean-Daniel Beaubien wrote: I am planning on running on test between maildir and mdbox to see which is a better fit for my use case. And I'm just looking for general advice/recommendation. I will post any results I obtain here. Maildir is good for reliability, since it's just about impossible to corrupt, and even in case of filesystem corruption it's easier to recover than other formats. mdbox is good if you want the best performance. Important question: I have multiple users hitting the same email account at the same time. Can be a problem with mdbox? No problem. - Serving only a hand full of email accounts but some of the accounst have over 3 millions emails in them (with individual mail folders having 100k+ emails) Maildir gets slow with that many mails in one folder. - fts-lucene or fts-solr? fts-lucene uses the latest CLucene version, which is a little old. With fts-solr you can use the latest Solr/Lucene. So as long as you don't mind setting up a Solr instance it should be better. The good thing about fts-lucene is that you can simply enable it and it works without any external servers.
Re: [Dovecot] maildir vs mdbox
On 28.1.2012, at 18.13, Jean-Daniel Beaubien wrote: Considering my use case (small number of accounts but alot of emails per account, and I should add that they are mostly small emails, most under 5k, alot under 30k) what mdbox setting would you recommend i start testing with (mdbox_rotate_size and mdbox_rotate_interval). mdbox_rotate_interval is useful only if you want smaller incremental backups (so files that are backed up no longer change unless messages are deleted). Its default is 0 (I just fixed example-config, which showed it as 1day). I don't really know about mdbox_rotate_size. It would be nice if someone were to test different values over longer period and report how it affects disk IO.
Re: [Dovecot] maildir vs mdbox
On Sat, Jan 28, 2012 at 11:37 AM, Timo Sirainen t...@iki.fi wrote: On 28.1.2012, at 18.13, Jean-Daniel Beaubien wrote: Considering my use case (small number of accounts but alot of emails per account, and I should add that they are mostly small emails, most under 5k, alot under 30k) what mdbox setting would you recommend i start testing with (mdbox_rotate_size and mdbox_rotate_interval). mdbox_rotate_interval is useful only if you want smaller incremental backups (so files that are backed up no longer change unless messages are deleted). Its default is 0 (I just fixed example-config, which showed it as 1day). To be honest, the smaller incremental backup part is interesting. That along with auto-gzip of the mdbox files are very interesting for me. I don't really know about mdbox_rotate_size. It would be nice if someone were to test different values over longer period and report how it affects disk IO. I was thinking on doing a test with 20MB and 80MB, look at the results and go from there. Btw, when I migrate my emails from Maildir to mdbox, dsync should take into account the rotate_size parameter. If I want to change the rotate_size parameter, I simply edit the config file, change the parameter (erase the mdbox folder?) and re-run dsync. Is that correct?
Re: [Dovecot] Crash on mail folder delete
On 28.1.2012, at 9.15, Daniel L. Miller wrote: Can I do anything to help find this? Folders are still shown in Trash - unable to delete. gdb backtrace would be helpful: http://dovecot.org/bugreport.html and doveconf -n and the folder name.
Re: [Dovecot] v2.1.rc5 (85a9b5236b6c) Error: lmtp client: DNS lookup of $FQDN failed: connect(dns-client) failed: No such file or directory
On 28.1.2012, at 17.34, Pascal Volk wrote: When the Sieve plugin tries to send a vacation message or redirect a message to another address it fails. dovecot: lmtp(6412, u...@example.com): Error: lmtp client: DNS lookup of orange.example.com failed: connect(dns-client) failed: No such file or directory Fixed: http://hg.dovecot.org/dovecot-2.1/rev/bc2eea348f55 http://hg.dovecot.org/dovecot-2.1/rev/32318f1588d4 The same problem exists in v2.0 also, but I didn't bother to fix it there. A workaround is to use IP instead of host in submission_host.
Re: [Dovecot] v2.1.rc5 (85a9b5236b6c) Error: lmtp client: DNS lookup of $FQDN failed: connect(dns-client) failed: No such file or directory
On 28.1.2012, at 19.29, Timo Sirainen wrote: On 28.1.2012, at 17.34, Pascal Volk wrote: When the Sieve plugin tries to send a vacation message or redirect a message to another address it fails. dovecot: lmtp(6412, u...@example.com): Error: lmtp client: DNS lookup of orange.example.com failed: connect(dns-client) failed: No such file or directory Fixed: http://hg.dovecot.org/dovecot-2.1/rev/bc2eea348f55 http://hg.dovecot.org/dovecot-2.1/rev/32318f1588d4 The same problem exists in v2.0 also, but I didn't bother to fix it there. A workaround is to use IP instead of host in submission_host. Oh, clarification: With LMTP it just happens to work with v2.0, but with dovecot-lda it doesn't work.
Re: [Dovecot] LMTP : Can't handle mixed proxy/non-proxy destinations
On 27.1.2012, at 12.59, Alexis Lelion wrote: Jan 25 09:05:12 mail01 postfix/lmtp[23934]: A92709300DB: to= user_on_mai...@domain.com, relay=mail01.domain.com[private/dovecot-lmtp], delay=0.07, delays=0.01/0/0/0.06, dsn=4.3.0, status=deferred (host mail01.domain.com[private/dovecot-lmtp] said: 451 4.3.0 user_on_mai...@domain.com Can't handle mixed proxy/non-proxy destinations (in reply to RCPT TO command)) I was wondering if there was another way of handling this, for example by triggering an immediate queue lookup from postfix or forwarding a copy of the mail to the other server. Note that the postfix queue_run_delay was increased to 15min on purpose, so I cannot change that. It would be possible to change the code to support mixed destinations, but it's probably not a simple change and I have other things to do.. Maybe you could work around it so that LMTP always proxies the mails, to localhost as well, but to a different port which doesn't do proxying at all.
Re: [Dovecot] IMAP to Maildir Migration preserving UIDs?
On 27.1.2012, at 2.00, Gedalya wrote: Starting program: /usr/bin/doveadm -o imapc_user=j...@example.com -o imapc_password= backup -u j...@example.com -R imapc: Program received signal SIGSEGV, Segmentation fault. mailbox_log_iter_open_next (iter=0x80cbd90) at mailbox-log.c:213 213mailbox-log.c: No such file or directory. in mailbox-log.c This crash is now fixed, so there's no need to give /tmp/imapc path anymore: http://hg.dovecot.org/dovecot-2.1/rev/7b94d1c8a6e7
Re: [Dovecot] Problem with Postfix + Dovecot + MySQL + Squirrelmail
On 27.1.2012, at 14.32, Gustavo wrote: #service dovecot start Starting IMAP/POP3 mail server: dovecotLast died with error (see error log for more information): Auth process died too early - shutting down No need to keep guessing the problem. See error log for more information like it says. http://wiki.dovecot.org/Logging
Re: [Dovecot] problem compiling imaptest under solaris
On 25.1.2012, at 16.43, Jürgen Obermann wrote: today I tried to compile imaptest under solaris 10 with studio 11 compiler and got the following error: gmake[2]: Entering directory `/net/fileserv/export/sunsrc/src/imaptest-2019/src' source='client.c' object='client.o' libtool=no \ DEPDIR=.deps depmode=none /bin/bash ../depcomp \ cc -DHAVE_CONFIG_H -I. -I. -I.. -I/opt/local/include/dovecot -I/usr/local/include -fast -xarch=v8plusa -I/usr/sfw/include -c client.c /opt/local/include/dovecot/imap-util.h, line 6: warning: useless declaration client-state.h, line 6: warning: useless declaration client.c, line 655: operand cannot have void type: op == client.c, line 655: operands have incompatible types: const void == int cc: acomp failed for client.c http://hg.dovecot.org/imaptest/rev/7e490e59f1ee should fix it?
Re: [Dovecot] Password auth scheme question with mysql
On 25.1.2012, at 2.03, Ed W wrote: The error seems to be that I set the pass variable in my password_query to set the master password for the upstream proxied to server. I can't actually remember now why this was required, but it was necessary to allow the proxy to work correctly in the past. I guess this assumption needs revisiting now since it can't be used if the plain password isn't in the database... I'm not sure if I understand correctly, but if you need the user's plaintext password it's in %w variable (assuming plaintext authentication). So a common configuration is to use: '%w' AS pass
Re: [Dovecot] [PATCH] autoconf small fix
On 24.1.2012, at 17.58, Luca Di Vizio wrote: the attached patch seems to solve a warning from autoconf: libtoolize: Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.in and libtoolize: rerunning libtoolize, to keep the correct libtool macros in-tree. I have considered it before, but I remember at one point there was some reason why I didn't want to do it. I just can't remember the reason anymore, maybe there isn't any.. But I don't really understand why libtoolize keeps complaining about that, since it works just fine as it is.
Re: [Dovecot] Quota-warning and setresgid
On 10.1.2012, at 11.34, l.chelchowski wrote: Jan 10 10:15:06 lda: Debug: auth input: tes...@domain.eu home=/home/vmail/domain.eu/tester/ mail=maildir:/home/vmail/domain.eu/tester/:INDEX=/var/mail/vmail/domain.eu/tes...@domain.eu/index/public uid=101 gid=12 quota_rule=*:storage=2097 acl_groups= Note that userdb lookup returns gid=12(mail) Jan 10 10:15:06 lda(tes...@domain.eu): Fatal: setresgid(12(mail),12(mail),101(vmail)) failed with euid=101(vmail): Operation not permitted But you're running it with gid=101(vmail). mail_gid = vmail mail_privileged_group = vmail mail_uid = vmail Here you're also using gid=101(vmail). (The mail_privileged_group=vmail is a useless setting BTW) userdb { args = /usr/local/etc/dovecot/dovecot-sql.conf driver = sql } My guess for the best fix: Change the user_query not to return uid or gid fields at all.
Re: [Dovecot] UID 0 problem while issuing an UID THREAD REFS command
On 23.1.2012, at 14.55, Jesús M. Navarro wrote: I'm having problems on a maildir due to dovecot returning an UID 0 to an UID THREAD REFS command: I'm sending to your personal address a whole maildir that reproduces the bug (it's very short) to avoid having it published in the mail archives. Thanks, I finally looked at this. The problem happens only when the THREADing isn't done for all messages. I thought this would have been a much more complex bug. Fixed: http://hg.dovecot.org/dovecot-2.0/rev/57498cad6ab9
Re: [Dovecot] where is subscribed list stored?
On 22.1.2012, at 15.55, Michael Makuch wrote: I use dovecot locally for internal only access to my email archives, of which I have many gigs of email archives. Over time I end up subscribing to a couple dozen different IMAP email folders. Problem is that periodically my list of subscribed folders get zapped to none, and I have to go and re-subscribe to a dozen or two folders again. Anyone seen this happen? It looks like the list of subscribed folders is here ~/Mail/.subscriptions and I can see in my daily backup that it reflects what appears in TBird. What might be zapping it? I use multiple email clients simultaneously on different hosts. (IOW I leave them open) Is this a problem? Does dovecot manage that in some way? Or is that my problem? I don't think this is the problem since this only occurs like a few times per year. If it were the problem I'd expect it to occur much more frequently. No idea, but you could prevent it by making sure that it can't change the subscriptions: mail_location = mbox:~/Mail:CONTROL=~/mail-subscriptions mkdir ~/mail-subscriptions mv ~/Mail/.subscriptions ~/mail-subscriptions chmod 0500 ~/mail-subscriptions I thought Dovecot would also log an error if client tried to change subscriptions, but looks like it doesn't. It only returns failure to client: a unsubscribe INBOX a NO [NOPERM] No permission to modify subscriptions
Re: [Dovecot] MySQL server has gone away
On 13.1.2012, at 20.29, Mark Moseley wrote: If there are multiple hosts, it seems like the most robust thing to do would be to exhaust the existing connections and if none of those succeed, then start a new connection to one of them. It will probably result in much more convoluted logic but it'd probably match better what people expect from a retry. Done: http://hg.dovecot.org/dovecot-2.0/rev/4e7676b890f1
Re: [Dovecot] imap-login process_limit reached
On 17.1.2012, at 16.23, Michael Orlitzky wrote: First of all, feature request: doveconf -d show the default value of all settings Done: http://hg.dovecot.org/dovecot-2.1/rev/41cb0217b7c3
Re: [Dovecot] 2.1.rc3 (1a722c7676bb):dsync umlaut problems
On 13.1.2012, at 5.10, Pascal Volk wrote: All umlauts in mailbox names are lost after converting mbox/Maildir mailboxes to mdbox. Looks like it was a generic problem in v2.1 dsync. Fixed: http://hg.dovecot.org/dovecot-2.1/rev/ef6f3b7f6038
Re: [Dovecot] moving mail out of alt storage
On 12.1.2012, at 20.32, Mark Moseley wrote: On Wed, 2011-09-14 at 23:17 -0400, Micah Anderson wrote: I moved some mail into the alt storage: doveadm altmove -u jo...@example.com seen savedbefore 1w and now I want to move it back to the regular INBOX, but I can't see how I can do that with either 'altmove' or 'mailbox move'. Is this sdbox or mdbox? With sdbox you could simply mv the files. Or apply patch: http://hg.dovecot.org/dovecot-2.0/rev/1910c76a6cc9 This is mdbox, which is why I am not sure how to operate because I am used to individual files as is with maildir. micah I'm curious about this too. Is moving the m.# file out of the ALT path's storage/ directory into the non-ALT storage/ directory sufficient? Or will that cause odd issues? You can manually move m.* files to alt storage and back. Just make sure that the same file isn't being simultaneously modified by Dovecot or you'll corrupt it.
Re: [Dovecot] dovecot 2.0.15 - purge errors
On 13.1.2012, at 8.20, Kamil Jońca wrote: Dovecot 2.0.15, debian package, am I lost some mails? How can I check what is in *.broken file? You can look at the .broken file with text editor for example :) --8---cut here---start-8--- $doveadm -v purge doveadm(kjonca): Error: Corrupted dbox file /home/kjonca/Mail/0/storage/m.6469 (around offset=291530): purging found mismatched offsets (291500 vs 299692, 60/215) 299692 - 291500 = 8192 = output stream's buffering size. I guess what happened is that sometimes earlier Dovecot crashed while it was saving a message, but it had managed to write 8192 bytes. Now purging notices the extra 8192 bytes and wonders what to do about them, so it starts index rebuild, which probably adds it as a new message to mailbox. In future this check probably should be done before appending the next message to mdbox, so it's noticed earlier and it probably should delete the message instead of adding a partially saved message to mailbox. doveadm(kjonca): Error: Corrupted dbox file /home/kjonca/Mail/0/storage/m.6469 (around offset=599914): metadata header has bad magic value This is about the same error as above. So, in short: Nothing to worry about. Although you could look into why the earlier saving crashed in the first place.
Re: [Dovecot] MySQL server has gone away
Am 28.01.2012 21:07, schrieb Timo Sirainen: On 13.1.2012, at 20.29, Mark Moseley wrote: If there are multiple hosts, it seems like the most robust thing to do would be to exhaust the existing connections and if none of those succeed, then start a new connection to one of them. It will probably result in much more convoluted logic but it'd probably match better what people expect from a retry. Done: http://hg.dovecot.org/dovecot-2.0/rev/4e7676b890f1 Hi Timo doc/example-config/dovecot-sql.conf.ext from hg has something like # Database connection string. This is driver-specific setting. # HA / round-robin load-balancing is supported by giving multiple host # settings, like: host=sql1.host.org host=sql2.host.org but i dont find it in http://wiki2.dovecot.org/AuthDatabase/SQL -- Best Regards MfG Robert Schetterer Germany/Munich/Bavaria
Re: [Dovecot] dovecot 2.0.15 - purge errors
On 28.1.2012, at 23.04, Timo Sirainen wrote: 299692 - 291500 = 8192 = output stream's buffering size. I guess what happened is that sometimes earlier Dovecot crashed while it was saving a message, but it had managed to write 8192 bytes. Now purging notices the extra 8192 bytes and wonders what to do about them, so it starts index rebuild, which probably adds it as a new message to mailbox. In future this check probably should be done before appending the next message to mdbox, so it's noticed earlier Done: http://hg.dovecot.org/dovecot-2.1/rev/bde005e302e0 and it probably should delete the message instead of adding a partially saved message to mailbox. Not done. Safer to not delete any data.
Re: [Dovecot] howto disable indexing on dovecot-lda ?
On 11.1.2012, at 19.30, Adrian Minta wrote: Hello, I tested with mail_location = whatever-you-have-now:INDEX=MEMORY and it seems to help, but in the mean time I found another option completely undocumented that seems to do exactly what I wanted: protocol lda { mailbox_list_index_disable= yes } Does anyone knows exactly what mailbox_list_index_disable does and if is still available in 2.0 and 2.1 branch ? mailbox_list_index_disable does absolutely nothing in v2.0, and it defaults to no in v2.1 also. It's about a different kind of index.
Re: [Dovecot] A small bug and a question about CONTEXT=SEARCH
On 11.1.2012, at 17.57, Anders wrote: Sorry, apparently I was a bit too fast there. ADDTO and REMOVEFROM should not be sent by a client, but I think that a client can send CONTEXT as a hint to the server, see http://tools.ietf.org/html/rfc5267#section-4.2 Yes, that was a bug. Thanks, fixed: http://hg.dovecot.org/dovecot-2.0/rev/fd16e200f0f7
Re: [Dovecot] sieve under lmtp using wrong homedir ?
On 11.1.2012, at 17.35, Frank Post wrote: All is working well except lmtp. Sieve scripts are correctly saved under /var/vmail/test.com/test/sieve, but under lmtp sieve will use /var/vmail//testuser/ Uid testuser has mail=t...@test.com configured in ldap. As i could see in the debug logs, there is a difference between the auth master out lines, but why ? .. Jan 11 14:39:53 auth: Debug: master in: USER1 testuser service=lmtplip=10.234.201.9rip=10.234.201.4 This means that Dovecot LMTP got: RCPT TO:testuser instead of: RCPT TO:t...@test.com You probably should fix your userdb lookup so that that would return unknown user instead of accepting it. But the real problem is anyway in your MTA setup.
Re: [Dovecot] MySQL server has gone away
On 28.1.2012, at 23.06, Robert Schetterer wrote: doc/example-config/dovecot-sql.conf.ext from hg has something like # Database connection string. This is driver-specific setting. # HA / round-robin load-balancing is supported by giving multiple host # settings, like: host=sql1.host.org host=sql2.host.org but i dont find it in http://wiki2.dovecot.org/AuthDatabase/SQL I added something about it there.
Re: [Dovecot] maildir vs mdbox
On 28.1.2012, at 19.02, Jean-Daniel Beaubien wrote: Btw, when I migrate my emails from Maildir to mdbox, dsync should take into account the rotate_size parameter. If I want to change the rotate_size parameter, I simply edit the config file, change the parameter (erase the mdbox folder?) and re-run dsync. Is that correct? Yes. You can also give -o mdbox_rotate_size=X parameter to dsync to override the config. The mdbox_rotate_size is started to be used immediately, so if you increase it Dovecot may start appending new mails to old files. The existing files aren't immediately shrunk, but during purge when writing new files the files can become smaller.
Re: [Dovecot] compressed mboxes very slow
I wonder if this patch helps here: http://hg.dovecot.org/dovecot-2.0/rev/9b2931607063 At least I can't now see any slowness with either v2.1 or the latest v2.0. But I don't know if I would have slowness with older versions either..
Re: [Dovecot] MySQL server has gone away
Am 28.01.2012 23:17, schrieb Timo Sirainen: On 28.1.2012, at 23.06, Robert Schetterer wrote: doc/example-config/dovecot-sql.conf.ext from hg has something like # Database connection string. This is driver-specific setting. # HA / round-robin load-balancing is supported by giving multiple host # settings, like: host=sql1.host.org host=sql2.host.org but i dont find it in http://wiki2.dovecot.org/AuthDatabase/SQL I added something about it there. cool thanks ! -- Best Regards MfG Robert Schetterer Germany/Munich/Bavaria
Re: [Dovecot] 2.0.17: Index lost - SAVEDON lost as well?
On 9.1.2012, at 16.57, Timo Sirainen wrote: After that, the SAVEDON date for all mails was reset to today: Yeah. The save date is stored only in index. And index rebuild drops all those fields. I guess this could/should be fixed in index rebuild. Fixed: http://hg.dovecot.org/dovecot-2.0/rev/c30ea8aec902
Re: [Dovecot] Attribute Cache flush errors on FreeBSD 8.2
On 7.1.2012, at 2.21, Phil Kernick wrote: I'm running dovecot 2.0.16 on FreeBSD 8.2 with the mail spool and indexes on an NFS server. Lines like the following keep appearing in syslog for access to each mailbox: Error: nfs_flush_attr_cache_fd_locked: fchown(/home/philk/Mail/Deleted) failed: Bad file descriptor I've given up on trying to make mail_nfs_* settings work. If you have only one Dovecot server, you don't need these settings at all. If you have more than one Dovecot server, use director (and then you also don't need these settings).
Re: [Dovecot] 2.1.rc1 (056934abd2ef): virtual plugin mailbox search pattern
On 23.12.2011, at 19.33, e-frog wrote: For testing propose I created the following folders with each containing one unread message INBOX, INBOX/level1 and INBOX/level1/level2 .. Result: virtual/unread shows only 1 unseen message. Further tests showed it's the one from INBOX. The mails from the deeper levels are not found. What mailbox format are you using? Maybe I fixed this with http://hg.dovecot.org/dovecot-2.1/rev/54e74090fb42
[Dovecot] Migration to multi-dbox and SiS
Dear list, A huge thank-you first of all for all the work that's gone into Dovecot itself. I'm rebuilding a mail server next week and so, taking the rare opportunity to re-consider all the options I've had running over the past couple of years. Around the time of the last re-build (2010), there had been some discussion on single instance storage, which was quite new on Dovecot around then. I chickened out of setting it up though. Now with it having been in the wild for a couple of years, I wonder, how have people found SiS to behave? Additionally, though there was talk of the prospect of it being merged with 2.x am I right in thinking it's not yet in the main project? Couldn't find any 2.x changelogs that could confirm this. With best wishes, Ronald.
Re: [Dovecot] Migration to multi-dbox and SiS
On 29.1.2012, at 1.16, Ronald MacDonald wrote: Around the time of the last re-build (2010), there had been some discussion on single instance storage, which was quite new on Dovecot around then. I chickened out of setting it up though. Now with it having been in the wild for a couple of years, I wonder, how have people found SiS to behave? Additionally, though there was talk of the prospect of it being merged with 2.x am I right in thinking it's not yet in the main project? Couldn't find any 2.x changelogs that could confirm this. It's in v2.0 and used by at least a few installations. Apparently it works quite well. As long as you have a pretty typical setup it should work fine. It gets more complex if you want to spread the data across multiple mount points. Backups may also be more difficult, since filesystem snapshots are pretty much the only 100% safe way to do them. BTW. SIS, not SiS (Instance, not in)
Re: [Dovecot] MySQL server has gone away
On Sat, Jan 28, 2012 at 12:07 PM, Timo Sirainen t...@iki.fi wrote: On 13.1.2012, at 20.29, Mark Moseley wrote: If there are multiple hosts, it seems like the most robust thing to do would be to exhaust the existing connections and if none of those succeed, then start a new connection to one of them. It will probably result in much more convoluted logic but it'd probably match better what people expect from a retry. Done: http://hg.dovecot.org/dovecot-2.0/rev/4e7676b890f1 Excellent, thanks!