- dovecot - dovecot.org

[Dovecot] Path for fixing sequence lookup of newly created mails in transaction view
by Potapov Sergey 19 Dec '12

19 Dec '12

Hi! It seem that tview_lookup_seq_range works incorret, if I request uids from 3 to 3, and I have new uids 1,2,4,5 it will return seq for uid=4. Fixing patch is included as attach. ------ Sergey Potapov

2 1

Re: [Dovecot] dovecot-lda (2.1.12) segfaults
by Joseph Tam 19 Dec '12

19 Dec '12

Timo Sirainen writes: > Can you still reproduce this in any way? Yes, I have 6 sets of user INBOX index caches that will crash dovecot-lda. The actual content of the INBOX is irrelevant (crash probably happens before INBOX is opened). > I found two bugs, would be nice to know if they solve it: > > http://hg.dovecot.org/dovecot-2.1/rev/2f848393f78e > http://hg.dovecot.org/dovecot-2.1/rev/bded819417d9 No, these patch don't help. It crashes in the same place with the same value of field_hdr. Here's the full backtrace: #0 0xff2a0474 in mail_cache_header_fields_read (cache=0x5c250) at mail-cache-fields.c:325 field_hdr = (const struct mail_cache_header_fields *) 0x20 field = {name = 0x402 <Address 0x402 out of bounds>, idx = 4282351288, type = MAIL_CACHE_FIELD_VARIABLE_SIZE, field_size = 4282335628, decision = MAIL_CACHE_DECISION_TEMP, last_used = -14558816} last_used = (const uint32_t *) 0x64584 sizes = (const uint32_t *) 0xc types = (const uint8_t *) 0x64888 "" decisions = (const uint8_t *) 0x64900 "" p = 0x24a38 "�\035\212@��" names = 0x0 end = 0x64a50 "" orig_key = (void *) 0xffbfee38 orig_value = (void *) 0x64550 fidx = 411784 new_fields_count = 4280126016 dec = MAIL_CACHE_DECISION_NO max_drop_time = 376804 offset = 32 i = 4282348464 #1 0xff29e8cc in mail_cache_compress_locked (cache=0x5c250, trans=0x645e0, unlock=0xffbfeeef) at mail-cache-compress.c:361 dotlock = (struct dotlock *) 0x2ea00 st = {st_dev = 235718347, st_pad1 = {874, 0, 0}, st_ino = 0, st_mode = 0, st_nlink = 0, st_uid = 0, st_gid = 0, st_rdev = 3720, st_pad2 = {0, 0}, st_size = 3720, st_atim = { tv_sec = 410816, tv_nsec = -12631336}, st_mtim = {tv_sec = 514, tv_nsec = -12631336}, st_ctim = {tv_sec = 65536, tv_nsec = 0}, st_blksize = 0, st_blocks = 1621028016851520, st_fstype = "\000\000\000\004\000\000\000\003\212\000\000\000\000\005�P", st_pad4 = {-4198784, -14028952, 39394339, 377424, 0, 16777216, 3, 4}} old_mask = 4282348464 file_seq = 4 old_offset = 4290768372 ext_offsets = {arr = {buffer = 0xffbfee10, element_size = 4280930288}, v = 0xffbfee10, v_modifiable = 0xffbfee10} offsets = (const uint32_t *) 0x0 data = (const void *) 0xff3f4380 i = 0 count = 1 fd = 0 ret = 377424 #2 0xff29efe0 in mail_cache_compress (cache=0x5c250, trans=0x645e0) at mail-cache-compress.c:489 unlock = false ret = 411764 __FUNCTION__ = "mail_cache_compress" #3 0xff2a3e28 in mail_cache_transaction_compress (ctx=0x5e3b8) at mail-cache-transaction.c:180 cache = (struct mail_cache *) 0x5c250 view = (struct mail_index_view *) 0x644c0 trans = (struct mail_index_transaction *) 0x645e0 ret = 2424 #4 0xff2a40b8 in mail_cache_transaction_open_if_needed (ctx=0x5e3b8) at mail-cache-transaction.c:241 cache = (struct mail_cache *) 0x5c250 ext = (const struct mail_index_ext *) 0x1e idx = 154968 i = 1 __FUNCTION__ = "mail_cache_transaction_open_if_needed" #5 0xff2a6e94 in mail_cache_field_want_add (ctx=0x5e3b8, seq=1, field_idx=12) at mail-cache-transaction.c:1048 decision = 153968 #6 0xff27e8e8 in index_mail_parse_header_register_all_wanted (mail=0x5efa8) at index-mail-headers.c:175 _mail = (struct mail *) 0x5efa8 all_cache_fields = (const struct mail_cache_field *) 0x25970 i = 12 count = 26 #7 0xff27ec90 in index_mail_parse_header_init (mail=0x5efa8, headers=0x0) at index-mail-headers.c:230 _data_stack_cur_id = 2 data = (struct index_mail_data *) 0x5f058 match = (const uint8_t *) 0x641a0 "" i = 0 field_idx = 4290769328 match_count = 2155905152 __FUNCTION__ = "index_mail_parse_header_init" #8 0xff27f5c8 in index_mail_cache_parse_init (_mail=0x5efa8, input=0x64058) at index-mail-headers.c:376 mail = (struct index_mail *) 0x5efa8 input2 = (struct istream *) 0x641a0 __FUNCTION__ = "index_mail_cache_parse_init" #9 0xff2299cc in mbox_save_get_input_stream (ctx=0x5e6e0, input=0x637c8) at mbox-save.c:411 filter = (struct istream *) 0x0 ret = (struct istream *) 0x5edd0 cache_input = (struct istream *) 0x25990 streams = {0x20202020, 0x2e938, 0xa202020} #10 0xff22a084 in mbox_save_begin (_ctx=0x5e6e0, input=0x637c8) at mbox-save.c:520 ctx = (struct mbox_save_context *) 0x5e6e0 t = (struct mbox_transaction_context *) 0x5de88 save_flags = MAIL_RECENT offset = 0 __FUNCTION__ = "mbox_save_begin" #11 0xff24e9c0 in mailbox_save_begin (ctx=0xffbff514, input=0x637c8) at mail-storage.c:1652 box = (struct mailbox *) 0x594e8 ret = 0 #12 0xff23f138 in mail_storage_try_copy (_ctx=0xffbff514, mail=0x54cd8) at mail-copy.c:68 ctx = (struct mail_save_context *) 0x5e6e0 pmail = (struct mail_private *) 0x54cd8 input = (struct istream *) 0x637c8 from_envelope = 0x13d90 "MAILER-DAEMON" guid = 0xff2f0ec0 "" received_date = -1 #13 0xff23f23c in mail_storage_copy (ctx=0x5e6e0, mail=0x54cd8) at mail-copy.c:93 No locals. #14 0xff24ec28 in mailbox_copy (_ctx=0xffbff670, mail=0x54cd8) at mail-storage.c:1721 ctx = (struct mail_save_context *) 0x5e6e0 box = (struct mailbox *) 0x594e8 keywords = (struct mail_keywords *) 0x0 ret = 389032 #15 0xff24ec98 in mailbox_save_using_mail (ctx=0xffbff670, mail=0x54cd8) at mail-storage.c:1730 No locals. #16 0xff388070 in mail_deliver_save (ctx=0xffbff8a8, mailbox=0x13fe8 "INBOX", flags=0, keywords=0x0, storage_r=0xffbff83c) at mail-deliver.c:317 open_ctx = {user = 0x3d3a8, lda_mailbox_autocreate = true, lda_mailbox_autosubscribe = false} box = (struct mailbox *) 0x594e8 trans_flags = MAILBOX_TRANSACTION_FLAG_EXTERNAL t = (struct mailbox_transaction_context *) 0x5de88 save_ctx = (struct mail_save_context *) 0x0 headers_ctx = (struct mailbox_header_lookup_ctx *) 0x0 kw = (struct mail_keywords *) 0x0 error = MAIL_ERROR_NONE mailbox_name = 0x13fe8 "INBOX" errstr = 0x0 guid = 0xff3f73b0 "" changes = {pool = 0x13e38, uid_validity = 0, saved_uids = {arr = {buffer = 0x13e28, element_size = 1}, v = 0x13e28, v_modifiable = 0x13e28}, ignored_modseq_changes = 4282350008, changed = false} range = (const struct seq_range *) 0xff1d3580 default_save = true ret = 0 __FUNCTION__ = "mail_deliver_save" #17 0xff38869c in mail_deliver (ctx=0xffbff8a8, storage_r=0xffbff83c) at mail-deliver.c:403 ret = -1 #18 0x00012d08 in main (argc=3, argv=0xffbff964) at main.c:434 set_roots = {0x24b48, 0x0} ctx = {pool = 0x2eaf0, set = 0x30440, session = 0x2eb00, dup_ctx = 0x0, session_id = 0x0, src_mail = 0x54cd8, src_envelope_sender = 0x0, dest_user = 0x3d3a8, dest_addr = 0x25828 "testuser@domain", final_dest_addr = 0x25828 "testuser@domain", dest_mailbox_name = 0x13fe8 "INBOX", dest_mail = 0x5efa8, var_expand_table = 0x0, tried_default_save = true, saved_mail = false, save_dest_mail = false, mailbox_full = false, dsn = false} service_flags = 1027 user = 0xffbffad0 "testuser" errstr = 0xff3f48e8 "" path = 0x0 storage_service = (struct mail_storage_service_ctx *) 0x2f650 service_user = (struct mail_storage_service_user *) 0x2fe88 service_input = {module = 0x13fd0 "lda", service = 0x13fd0 "lda", username = 0xffbffad0 "testuser", session_id = 0x0, local_ip = {family = 0, u = {ip6 = { _S6_un = {_S6_u8 = '\0' <repeats 15 times>, _S6_u32 = {0, 0, 0, 0}, __S6_align = 0}}, ip4 = {S_un = {S_un_b = {s_b1 = 0 '\0', s_b2 = 0 '\0', s_b3 = 0 '\0', s_b4 = 0 '\0'}, S_un_w = {s_w1 = 0, s_w2 = 0}, S_addr = 0}}}}, remote_ip = {family = 0, u = {ip6 = { _S6_un = {_S6_u8 = '\0' <repeats 15 times>, _S6_u32 = {0, 0, 0, 0}, __S6_align = 0}}, ip4 = {S_un = {S_un_b = {s_b1 = 0 '\0', s_b2 = 0 '\0', s_b3 = 0 '\0', s_b4 = 0 '\0'}, S_un_w = {s_w1 = 0, s_w2 = 0}, S_addr = 0}}}}, local_port = 0, remote_port = 0, userdb_fields = 0x0, flags_override_add = 0, flags_override_remove = 0, no_userdb_lookup = 0} storage = (struct mail_storage *) 0x39330 user_source = 0x13f30 "" destaddr_source = 0x13f30 "" process_euid = 0 stderr_rejection = false ret = 1 c = -1 error = MAIL_ERROR_NONE Joseph Tam <tam(a)math.ubc.ca>

2 2

[Dovecot] Too many imap connections in state "idling"
by 3.listas＠adminlinux.com.br 19 Dec '12

19 Dec '12

Hi, I have a "Ubuntu12.04 + dovecot-2.0.19" configuration in my server. My mailbox server is shared by ~ 10k domains. It works fine with ~50k accounts. On average, my server has about 6K simultaneous imap connections during office hours. But I've noticed that this server has an average of approximately 4k connections in state "idling". On other Dovecot servers the average connections in state "idling" is only 10%. My doveconf -n: http://thiagoh.prv.digirati.com.br/doveconf-n.txt Can someone help me? Thanks -- Thiago Henrique adminlinux.com.br

4 5

[Dovecot] Dsync help, new attempt
by Torbjorn Astlind 19 Dec '12

19 Dec '12

I observe that dsync fills Maildir correctly incl. cur, if I delete Maildir just before dsync runs. Can the failed mbox to Maildir conversion be caused by dovecot creating Maildir automatically if a client connects to dovecot? Should dovecot be shutdown when running dsync? All the best, Torbjörn 9. dsync help? (Torbjorn Astlind) ------------------------------ Message: 9 Date: Mon, 10 Dec 2012 16:33:08 +0100 From: Torbjorn Astlind <torbjorn(a)dbb.su.se> To: dovecot(a)dovecot.org Subject: [Dovecot] dsync help? Message-ID: <50C600B4.4040603(a)dbb.su.se> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Hi all, I have dovecot serving aprox 200 users through mbox, but the time has come to migrate to a new server, and to use Maildir. Using rsync from the old server I get all user data, i.e /home/*, and inboxes i.e /var/mail/* to a new server. This has worked well for some time as backup. The dsync should convert everything to Maildir, and it does... to an extent. (linebrakes for easy viewing) What I see, is that often, the /home/mail/xxx is NOT resulting in /home/xxx/Maildir/cur All mailboxes from /home/xxx, are properly converted to /home/xxx/Maildir/.name_of_ box Am I misunderstanding something? I get it to work sometimes if I delete the ~/Maildir completely before doing the dsync. All, and any help most welcome, Torbj?rn for i in `cat /home2/torbjorn/SCRIPTS/userlist.txt`; do dsync -D -v -f -u $i -c /etc/dovecot/alt_config/10-mail.MBOX.conf backup maildir:~/Maildir; done > /home2/torbjorn/SCRIPTS/mb2mdRES.txt 2>&1 the -c 10.mail.MBOX.conf contains, the parts I think are relevant for the dsync operation: mail_location = mbox:~/home:INBOX=/home/mail/%u namespace { type = private separator = '/' inbox = yes } dsync(root): Debug: Effective uid=501, gid=501, home=/home/torbjorn dsync(root): Debug: Namespace : type=private, prefix=, sep=/, inbox=yes, hidden=no, list=yes, subscriptions=yes location=mbox:~/home:INBOX=/home/mail/torbjorn dsync(root): Debug: fs: root=/home/torbjorn/home, index=, control=, inbox=/home/mail/torbjorn, alt= dsync(root): Debug: Effective uid=501, gid=501, home=/home/torbjorn dsync(root): Debug: Namespace : type=private, prefix=, sep=/, inbox=yes, hidden=no, list=yes, subscriptions=yes location=maildir:~/Maildir dsync(root): Debug: maildir++: root=/home/torbjorn/Maildir, index=, control=, inbox=/home/torbjorn/Maildir, alt= dsync(root): Debug: Namespace : /home/torbjorn/Maildir doesn't exist yet, using default permissions dsync(root): Debug: Namespace : Using permissions from /home/torbjorn/Maildir: mode=0700 gid=-1 dsync(torbjorn): Debug: Namespace : Using permissions from /home/torbjorn/home: mode=0700 gid=-1 ------------------------------ _______________________________________________ dovecot mailing list dovecot(a)dovecot.org http://dovecot.org/cgi-bin/mailman/listinfo/dovecot End of dovecot Digest, Vol 116, Issue 16 ****************************************

1 0

[Dovecot] Advice on sieve vacation for aliases
by Marcio Merlone 19 Dec '12

19 Dec '12

Greetings, I have a working mail server with sieve 1:1.2.9-1ubuntu6.5 working just fine. My commercial dept. has asked me to set a vacation message for comercial(a)domain.tld, but this is an alias to jhon(a)domain.tld and meggie(a)domain.tld. I have created a sieve_global_path = /etc/dovecot/global.sieve and it works, but not the intended way: it sends two vacation messages, each From: each employee, and not From: comercial(a)domain.tld alias. Is there how to make dovecot send only one vacation message From: comercial(a)domain.tld or will I have to survive with one message from jhon and another from meggie? Also, is there how to have a dovecot.lda-dupes for the alias instead the users? /etc/dovecot/global.sieve: require ["vacation"]; # rule:[Ferias coletivas] if anyof (header :contains "To" "comercial(a)domain.tld") { vacation :days 7 :subject "Férias coletivas de 20/12/2012 a 07/01/2013" text: bla bla bla. ; stop; } Thanks in advance and nice holidays. -- Marcio Merlone

2 1

[Dovecot] Adding a parameter to dovecot's deliver to get a meaningful return value
by Cyril RUSSO 18 Dec '12

18 Dec '12

Hi, First, please forgive me as I'm not subscribed to the list. I'm using Dovecot 2 (2.1.7) for serving my local maildir. This maildir is filled by an external process, fetching & merging mails from different sources that's outside this scope (but it's working). Dovecot is thus acting as a MTA, but more as a MDA. I've a web client running on the server, and it connects to Dovecot to get the mails. I've configured this web client & dovecot to use sieve to filter the mails. I know that since Dovecot does not actually deliver the messages, it's not the expected behavior. Anyway, I can use the "deliver" software in your suite for passing sieve's rules on each message in my maildir, and it's almost working. I've a cron job that run deliver on each message of my maildir, but I'm a bit puzzled now. Basically, I'm not able to figure out if a message was matched by a sieve filter or not. I'm a bit disappointed by two things: 1) Sieve's reject doesn't make "deliver" to return an error, so a sieve rule chain ending in "reject" can't be detected from my cron's script. I wonder it's due to the step deliver is supposed to be run (after a mail is already accepted). 2) "deliver" doesn't have a method to find out if it processed the given mail as input. Basically, would it be possible to either add a parameter to "deliver" (for example: "--fail-on-discarding") so it would fail if the sieve rules discard or reject (/ereject) the mail ? That way, I could detect the failure in my script, and know that I don't need to delete the initial mail since it was not filtered. Or, thought differently, I wonder if "deliver" could open a socket/pipe/file whatever, and write the number of processed mails in there (or using a "--return-processed-mail"), so a sieve rule ending with "delete", would be detectable by my script. Best regards, -- Cyril

2 1

Re: [Dovecot] Undeserved permissions error with g+s folder permissions
by tlhackque 18 Dec '12

18 Dec '12

> It's an MBox file that I'm trying to rename, not a directory. > | file Vendors/|| > ||Vendors/: setgid directory|| > || file Vendors/Test|| > ||Vendors/Test: empty|| > | > (Perhaps you missed that I said I created it "messages only", and the > config I originally posted showed the file storage is mbox...) > > So perhaps dovecot is confused - expecting an mbox file to have > directory permissions, when the directory it's in is setgid? At least > that's how it has seemed to me all along... > > Thanks. > > This communication may not represent my employer's views, > if any, on the matters discussed. > On 18-Dec-12 14:28, Reindl Harald wrote: >> Am 18.12.2012 19:39, schrieb Timo Sirainen: >>> On Sat, 2012-12-01 at 10:59 -0500, tlhackque wrote: >>>> It contains a subdirectory: >>>> >>>> -rw-rw---- 1 thlackque mail 84805345 Dec 1 10:34 Vendors/AVendor >>>> >>>> If I try to rename the AVendor folder (client is Thunderbird 17.0), I >>>> get this error: >>>> >>>> Unable to rename across conflicting directory permisssions >> this is 660 which is simply wrong for a directrory >> in case of diretcories you need the excue flag means 770 >> > -- This communication may not represent my employer's views, if any, on the matters discussed.

1 0

[Dovecot] Clear-text ldap passwords: {CLEAR} scheme?
by Kristof Bajnok 18 Dec '12

18 Dec '12

Hi, if I read the documentation correctly, the clear-text password storage scheme in Dovecot is called PLAIN. Unfortunately, in Sun LDAP Directory Server I have to specify them as "{CLEAR}" [1]. Is it somehow possible to make Dovecot recognise "{clear}foobar" passwords as plaintext? Now I get: auth: Error: ldap(username): Unknown scheme clear Thanks, Kristof [1]: http://docs.oracle.com/cd/E19850-01/816-6699-10/confattr.html#passwordStora…

2 1

[Dovecot] dovecot-lda (2.1.12) segfaults
by Joseph Tam 18 Dec '12

18 Dec '12

We uograded our dovecot from version 2.1.10 -> 2.10.12, but within the first hour of use, dovecot-lda would segfault during delivery. This left a lock file lying around causing the user's mail readers to hang; much hilarity ensues. All further deliveries to the same user will result in a crash. The only thing that fixes this condition is to blow away the user's INBOX cache index. Reindexing with "doveadm index" does not fix anything. I've installed a test version with one user's problematic INBOX cache indices and ran gdb dovecot-lda (gdb) run -t testuser </dev/null and it crashes with the following backtrace Program received signal SIGSEGV, Segmentation fault. 0xff2a0440 in mail_cache_header_fields_read (cache=0x5c120) at mail-cache-fields.c:323 323 if (field_hdr->size < sizeof(*field_hdr) + (gdb) bt #0 0xff2a0440 in mail_cache_header_fields_read (cache=0x5c120) at mail-cache-fields.c:323 #1 0xff29e8bc in mail_cache_compress_locked (cache=0x5c120, trans=0x63ce8, unlock=0xffbfeeef) at mail-cache-compress.c:361 #2 0xff29efd0 in mail_cache_compress (cache=0x5c120, trans=0x63ce8) at mail-cache-compress.c:489 #3 0xff2a3df4 in mail_cache_transaction_compress (ctx=0x5dac0) at mail-cache-transaction.c:180 #4 0xff2a4084 in mail_cache_transaction_open_if_needed (ctx=0x5dac0) at mail-cache-transaction.c:241 #5 0xff2a6e60 in mail_cache_field_want_add (ctx=0x5dac0, seq=1, field_idx=12) at mail-cache-transaction.c:1048 #6 0xff27e8e8 in index_mail_parse_header_register_all_wanted (mail=0x5e6b0) at index-mail-headers.c:175 #7 0xff27ec90 in index_mail_parse_header_init (mail=0x5e6b0, headers=0x0) at index-mail-headers.c:230 #8 0xff27f5c8 in index_mail_cache_parse_init (_mail=0x5e6b0, input=0x63760) at index-mail-headers.c:376 #9 0xff2299cc in mbox_save_get_input_stream (ctx=0x5dde8, input=0x62ed0) at mbox-save.c:411 #10 0xff22a084 in mbox_save_begin (_ctx=0x5dde8, input=0x62ed0) at mbox-save.c:520 #11 0xff24e9c0 in mailbox_save_begin (ctx=0xffbff514, input=0x62ed0) at mail-storage.c:1652 #12 0xff23f138 in mail_storage_try_copy (_ctx=0xffbff514, mail=0x54cd8) at mail-copy.c:68 #13 0xff23f23c in mail_storage_copy (ctx=0x5dde8, mail=0x54cd8) at mail-copy.c:93 #14 0xff24ec28 in mailbox_copy (_ctx=0xffbff670, mail=0x54cd8) at mail-storage.c:1721 #15 0xff24ec98 in mailbox_save_using_mail (ctx=0xffbff670, mail=0x54cd8) at mail-storage.c:1730 #16 0xff388070 in mail_deliver_save (ctx=0xffbff8a8, mailbox=0x13fe8 "INBOX", flags=0, keywords=0x0, storage_r=0xffbff83c) at mail-deliver.c:317 #17 0xff38869c in mail_deliver (ctx=0xffbff8a8, storage_r=0xffbff83c) at mail-deliver.c:403 #18 0x00012d08 in main (argc=3, argv=0xffbff964) at main.c:434 I dumped the value of field_hdr, and it has some bogus value (gdb) p field_hdr $1 = (const struct mail_cache_header_fields *) 0x20 Other observations - blowing away the indices does not fix it permanantly. The the LDA will eventually start crashing again. - the problematic indices do not cause problems with 2.1.10. Reverting to 2.1.10 restores proper mail delivery. Is this enough debug info to work your magic, Timo? I can supply the cache data if you neded it. It's compiled on a Solaris10. Joseph Tam <jtam.home(a)gmail.com>

2 1

[Dovecot] NFS, deleting folder issue
by Catalin Petrescu 18 Dec '12

18 Dec '12

Hi there, I use dovecot on a server with some Nightmare File Systerm (aka NFS) storage. When I delete a folder, its content gets deleted, but the folder itself does not. Here's what I found: #### Step 1: From a shell, telnet to imap, then create a folder cpetrescu@cpetrescu-mobile:~$ telnet server1.localdomain 143 Trying 10.200.28.36... Connected to server1.localdomain. Escape character is '^]'. * OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE AUTH=PLAIN AUTH=LOGIN] Dovecot ready. 01 login testuser1 Password 01 OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE SORT SORT=DISPLAY THREAD=REFERENCES THREAD=REFS MULTIAPPEND UNSELECT CHILDREN NAMESPACE UIDPLUS LIST-EXTENDED I18NLEVEL=1 CONDSTORE QRESYNC ESEARCH ESORT SEARCHRES WITHIN CONTEXT=SEARCH LIST-STATUS] Logged in 02 create INBOX/F1 02 OK Create completed. #### Step2: On the server, everything is OK, the folder is there [server1: testuser1]# ls -lrtA Maildir/INBOX/ total 4 drwxr-xr-x 5 root root 4096 Dec 5 10:28 F1 [server1: testuser1]# ls -lrtA Maildir/INBOX/F1/ total 12 drwxr-xr-x 2 root root 4096 Dec 5 10:28 tmp drwxr-xr-x 2 root root 4096 Dec 5 10:28 new drwxr-xr-x 2 root root 4096 Dec 5 10:28 cur #### Step3: in the telnet session, delete the folder 03 delete INBOX/F1 03 OK Delete completed. #### Step4: On the server, the content of the folder is gone, but the folder is still there [server1: testuser1]# ls -lrtA Maildir/INBOX/ total 4 drwxr-xr-x 2 root root 4096 Dec 5 10:29 F1 [server1: testuser1]# ls -lrtA Maildir/INBOX/F1/ total 0 #### Debugging: I defined my own my_delete_dir and added it to the delete_dir list in mailbox_list_vfuncs In my_delete_dir I just call readdir on that folder, and print the files via syslog. Here's the content of F1 before calling delete_dir: Dec 05 10:29:14 imap(testuser1): my_delete_dir: d_name: "new" inode: 843617 Dec 05 10:29:14 imap(testuser1): my_delete_dir: d_name: "tmp" inode: 843623 Dec 05 10:29:14 imap(testuser1): my_delete_dir: d_name: "cur" inode: 843616 Dec 05 10:29:14 imap(testuser1): my_delete_dir: d_name: "dovecot.index.log" inode: 2363059 Dec 05 10:29:14 imap(testuser1): my_delete_dir: d_name: "." inode: 843612 Dec 05 10:29:14 imap(testuser1): my_delete_dir: d_name: ".." inode: 807717 And here's its content after calling delete_dir Dec 05 10:29:15 imap(testuser1): my_delete_dir: name: INBOX/F1 Dec 05 10:29:15 imap(testuser1): my_delete_dir: path: /data/testuser1/Maildir/INBOX/F1 Dec 05 10:29:15 imap(testuser1): my_delete_dir: ret: -1 Dec 05 10:29:15 imap(testuser1): my_delete_dir: error: "Directory not empty" trying again Dec 05 10:29:15 imap(testuser1): my_delete_dir: d_name: "." inode: 843612 Dec 05 10:29:15 imap(testuser1): my_delete_dir: d_name: ".." inode: 807717 Dec 05 10:29:15 imap(testuser1): my_delete_dir: d_name: ".nfs0000000000240eb3000000e0" inode: 2363059 Dec 05 10:29:15 imap(testuser1): my_delete_dir: ret: -1 So, the dovecot.index.log was deleted, but it's still open, and the folder can't be deleted because of the pesky .nfs* file. Now, the question: Is there a *magic* function that closes "dovecot.index.log" that I can call before delete_dir()? I tried mailbox_log_close() but it didn't help. Thanks a lot. p.s. Please don't suggest that I should not use NFS! I know that, and I don't really like NFS, but I have no other choice. Catalin Petrescu <catalin.petrescu gmail com>

2 2

[Dovecot] SPECIAL-USE and mailbox names in various IMAP clients
by Timo Sirainen 18 Dec '12

18 Dec '12

http://imapwiki.org/SpecialUse Please fill out for all clients with all languages you know of. Knowing the mailbox names can be helpful with Dovecot's mailbox_alias plugin.

2 2

[Dovecot] Manually creating dovecot- files after isync/mbsync mirror
by Dale Gallagher 18 Dec '12

18 Dec '12

Hi everyone I'm using isync 1.0.5 patched to do recursive folder listing and mirroring mail from an external party's IMAP server to a local Maildir, which will be accessed by Dovecot 2.0.9. In order to prevent IMAP clients from re-downloading their mail, when connecting to Dovecot, I'm busy writing a script to generate appropriate dovecot-uidlist and dovecot-uidvalidity files, per Maildir. isync stores the uid in the filename, so that's extractable, per message, and the uidvalidity and current max uid are stored in a file .uidvalidity, in each Maildir. I've read the docs here: http://wiki2.dovecot.org/MailboxFormat/Maildir Regarding the dovecot-uidlist header (first line) - I've tried to search for and figure out what exactly is meant by the "mailbox global uid" and once I do, how to manually generate the 128 bit hex variation of it? Could someone elaborate on this, please. How would omitting the global uid field, beginning with G, affect remote clients' syncing of mail? Is this field mandatory? I also see that Dovecot's uidvalidity is of the form 4df8a4e2, whereas isync uses an integer. I'd appreciate it if someone could elaborate on what Dovecot uses and whether I can convert one to the other? I'll post the full solution, once complete, and working. Thanks. Dale

2 1

Re: [Dovecot] Automatically Cleaning Kerberos Credential Cache Files
by Joseph Tam 18 Dec '12

18 Dec '12

Ben Morrow writes: > In any case, it's likely that the delivery process looks in the user's > home directory even if delivery is to a separate mail spool, unless > you've taken steps to prevent this. I can confirm this. I think the reason is that dovecot changes to the user's home directory so that it has a place to dump core. This causes problem for my setup since I user per-user NFS mounts, so a delivery to an internal mailing list caused hundreds of NFS automounts. I patched this behaviour out, but there is probably a more elegant solution (unset MAIL_STORAGE_SERVICE_FLAG_ENABLE_CORE_DUMPS? How?). Joseph Tam <jtam.home(a)gmail.com> --- ./src/lib-storage/mail-storage-service.c~ Sat May 19 08:42:12 2012 +++ ./src/lib-storage/mail-storage-service.c Mon Jun 25 22:00:57 2012 @@ -615,21 +615,21 @@ - if ((user->flags & MAIL_STORAGE_SERVICE_FLAG_TEMP_PRIV_DROP) != 0 && - (user->flags & MAIL_STORAGE_SERVICE_FLAG_ENABLE_CORE_DUMPS) == 0) { - /* we don't want to write core files to any users' home - directories since they could contain information about other - users' mails as well. so do no chdiring to home. */ - } else if (*home != '\0' && - (user->flags & MAIL_STORAGE_SERVICE_FLAG_NO_CHDIR) == 0) { - /* If possible chdir to home directory, so that core file - could be written in case we crash. */ - if (chdir(home) < 0) { - if (errno == EACCES) { - i_error("%s", eacces_error_get("chdir", - t_strconcat(home, "/", NULL))); - } if (errno != ENOENT) - i_error("chdir(%s) failed: %m", home); - else if (mail_set->mail_debug) - i_debug("Home dir not found: %s", home); - } - } +// if ((user->flags & MAIL_STORAGE_SERVICE_FLAG_TEMP_PRIV_DROP) != 0 && +// (user->flags & MAIL_STORAGE_SERVICE_FLAG_ENABLE_CORE_DUMPS) == 0) { +// /* we don't want to write core files to any users' home +// directories since they could contain information about other +// users' mails as well. so do no chdiring to home. */ +// } else if (*home != '\0' && +// (user->flags & MAIL_STORAGE_SERVICE_FLAG_NO_CHDIR) == 0) { +// /* If possible chdir to home directory, so that core file +// could be written in case we crash. */ +// if (chdir(home) < 0) { +// if (errno == EACCES) { +// i_error("%s", eacces_error_get("chdir", +// t_strconcat(home, "/", NULL))); +// } if (errno != ENOENT) +// i_error("chdir(%s) failed: %m", home); +// else if (mail_set->mail_debug) +// i_debug("Home dir not found: %s", home); +// } +// }

2 1

[Dovecot] Removing configured SSL certificate file from disk while Dovecot is running spams syslog
by Nils 18 Dec '12

18 Dec '12

Hi! [root@mikan ~]# dovecot --version 2.1.12 When I remove the SSL certificate from disk while Dovecot is running, it spams syslog with the following log message over and over: Dec 6 17:41:41 mikan dovecot: config: Fatal: Error in configuration file /etc/dovecot/conf.d/10-ssl.conf line 12: ssl_cert: Can't open file /etc/ssl/private/dovecot.pem: No such file or directory Dec 6 17:41:41 mikan dovecot: config: Fatal: Error in configuration file /etc/dovecot/conf.d/10-ssl.conf line 12: ssl_cert: Can't open file /etc/ssl/private/dovecot.pem: No such file or directory Dec 6 17:41:41 mikan dovecot: config: Fatal: Error in configuration file /etc/dovecot/conf.d/10-ssl.conf line 12: ssl_cert: Can't open file /etc/ssl/private/dovecot.pem: No such file or directory [...] Reporting that the certificate went away is fine and useful, but spamming the syslog over and over with hundreds of entries per second(!) is annoying. :( I know what I'm doing, so reporting it once per connection attempt or something is enough. Even if Dovecot just decides to terminate, that would be fine. But a denial of service attack on syslog is not nice. Configuration output follows below. All the best, Nils [root@mikan ~]# dovecot -n # 2.1.12: /etc/dovecot/dovecot.conf # OS: Linux 3.5.4-1-ARCH x86_64 auth_failure_delay = 5 secs auth_mechanisms = plain login auth_verbose = yes listen = *,:: login_greeting = mail_location = maildir:~/.maildir namespace inbox { inbox = yes location = mailbox Drafts { special_use = \Drafts } mailbox Junk { special_use = \Junk } mailbox Sent { special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } mailbox Trash { special_use = \Trash } prefix = } passdb { driver = pam } protocols = pop3 service pop3-login { inet_listener pop3 { port = 110 } inet_listener pop3s { port = 995 ssl = yes } } service pop3 { process_limit = 10 } ssl_cert = </etc/ssl/private/dovecot.pem ssl_key = </etc/ssl/private/dovecot.key userdb { driver = passwd } protocol imap { imap_idle_notify_interval = 2 mins mail_max_userip_connections = 5 } protocol pop3 { mail_max_userip_connections = 5 pop3_logout_format = pop3_save_uidl = no pop3_uidl_format = %08Xu%08Xv }

2 1

[Dovecot] During FTS reindexing, search commands unceremoniously exit
by Dave Abrahams 18 Dec '12

18 Dec '12

Is this intentional? Closing the connection without warning doesn't seem like the best response... --8<---------------cut here---------------start------------->8--- % printf "1 select \"INBOX\"\n2 search body quack\n3 logout\n" | /opt/local/libexec/dovecot/imap * PREAUTH [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE SORT SORT=DISPLAY THREAD=REFERENCES THREAD=REFS THREAD=ORDEREDSUBJECT MULTIAPPEND UNSELECT CHILDREN NAMESPACE UIDPLUS LIST-EXTENDED I18NLEVEL=1 CONDSTORE QRESYNC ESEARCH ESORT SEARCHRES WITHIN CONTEXT=SEARCH LIST-STATUS SPECIAL-USE SEARCH=FUZZY] Logged in as dave * FLAGS (\Answered \Flagged \Deleted \Seen \Draft gnus-forward gnus-expire) * OK [PERMANENTFLAGS (\Answered \Flagged \Deleted \Seen \Draft gnus-forward gnus-expire \*)] Flags permitted. * 31 EXISTS * 0 RECENT * OK [UIDVALIDITY 1349969399] UIDs valid * OK [UIDNEXT 1376] Predicted next UID * OK [HIGHESTMODSEQ 2908] Highest 1 OK [READ-WRITE] Select completed. imap(dave): Info: Connection closed in=45 out=2027404 % --8<---------------cut here---------------end--------------->8--- -- Dave Abrahams BoostPro Computing Software Development Training http://www.boostpro.com Clang/LLVM/EDG Compilers C++ Boost

2 1

[Dovecot] Documentation of Redis and Memcache Backends
by Urban Loesch 18 Dec '12

18 Dec '12

Hi, in the release notes of 2.1.9 I read that dovecot supports memcache and redis backends for userdb/passdb authentication. This is very interesting for me and should reduce queries and database load of our mysql servers. My idea is to use memcache or redis in our IMAP/POP3 proxies in front of our backend servers. So I would like to try out if it's possible to store proxy information of our backends. For example the backend ip adress. But in the wiki I found only a few configuration settings for redis backend. http://master.wiki2.dovecot.org/AuthDatabase/Dict Also the mentioned example config file "dovecot-dict-auth.conf.ext" with full list of configuration options does not exits in the source of 2.1.11. Have you any idea where I can find the full info or any howtos? Many thanks and regards Urban

2 1

[Dovecot] Error: unlink_directory(.../lucene-indexes) with doveadm index
by Fred Kilbourn 18 Dec '12

18 Dec '12

Hi, Per discussion on this previous thread: http://dovecot.org/pipermail/dovecot/2012-October/068849.html I am updating search indexes on all my mailboxes via a nightly cronjob that runs: doveadm index -A '*' The problem we're running into is that random mailboxes will throw the following error during the update (sanitized to conceal actual domain name): doveadm(user(a)domain.com): Error: unlink_directory(/var/spool/mail/domain.com/user/sdbox/lucene-indexes) failed: Permission denied We have around 250 domains on the server, with a handful of mailboxes on average in each domain. We get this for a varying number of total mailboxes (1 to 6 every night). It is common for the same mailbox to show up a few days in a row and then stop throwing the error. The script is running as root, so there shouldn't be any file level permissions. Any thoughts on the root cause of these errors? What can I provide to help further diagnose? Thanks, Fred Kilbourn

2 2

[Dovecot] indexer-worker Panic
by Dave Abrahams 18 Dec '12

18 Dec '12

I just found this in my logs. Any idea? --8<---------------cut here---------------start------------->8--- 12/4/12 12:33:43.831 PM dovecot[2660]: indexer-worker(dave): Panic: file mailbox-list-fs.c: line 156 (fs_list_get_path): assertion failed: (mailbox_list_is_valid_pattern(_list, name)) 12/4/12 12:33:43.831 PM dovecot[2660]: indexer-worker(dave): Error: Raw backtrace: 2 libdovecot.0.dylib 0x000000010840e6c4 default_fatal_finish + 36 -> 3 libdovecot.0.dylib 0x000000010840f3e1 i_internal_fatal_handler + 49 -> 4 libdovecot.0.dylib 0x000000010840e98e i_panic + 158 -> 5 libdovecot-storage.0.dylib 0x0000000108340693 fs_list_get_path + 467 -> 6 indexer-worker 0x00000001082e7341 master_connection_input + 513 -> 7 libdovecot.0.dylib 0x000000010841b5 12/4/12 12:33:43.831 PM dovecot[2660]: indexer-worker(dave): Error: ce io_loop_call_io + 46 -> 8 libdovecot.0.dylib 0x000000010841c737 io_loop_handler_run + 215 -> 9 libdovecot.0.dylib 0x000000010841b7e8 io_loop_run + 88 -> 10 libdovecot.0.dylib 0x00000001084079c8 master_service_run + 24 -> 11 indexer-worker 0x00000001082e7001 main + 321 -> 12 libdyld.dylib 0x00007fff93d117e1 start + 0 -> 13 ??? 0x0000000000000001 0x0 + 1 12/4/12 12:33:44.035 PM dovecot[2660]: indexer: Error: Indexer worker disconnected, discarding 5 requests for dave 12/4/12 12:33:44.036 PM dovecot[2660]: indexer-worker(dave): Fatal: master: service(indexer-worker): child 2665 killed with signal 6 (core dumps disabled) 12/4/12 12:34:17.094 PM dovecot[2660]: indexer-worker(dave): Panic: file mailbox-list-fs.c: line 156 (fs_list_get_path): assertion failed: (mailbox_list_is_valid_pattern(_list, name)) 12/4/12 12:34:17.094 PM dovecot[2660]: indexer-worker(dave): Error: Raw backtrace: 2 libdovecot.0.dylib 0x000000010d1dd6c4 default_fatal_finish + 36 -> 3 libdovecot.0.dylib 0x000000010d1de3e1 i_internal_fatal_handler + 49 -> 4 libdovecot.0.dylib 0x000000010d1dd98e i_panic + 158 -> 5 libdovecot-storage.0.dylib 0x000000010d10d693 fs_list_get_path + 467 -> 6 indexer-worker 0x000000010d0b7341 master_connection_input + 513 -> 7 libdovecot.0.dylib 0x000000010d1ea5 12/4/12 12:34:17.094 PM dovecot[2660]: indexer-worker(dave): Error: ce io_loop_call_io + 46 -> 8 libdovecot.0.dylib 0x000000010d1eb737 io_loop_handler_run + 215 -> 9 libdovecot.0.dylib 0x000000010d1ea7e8 io_loop_run + 88 -> 10 libdovecot.0.dylib 0x000000010d1d69c8 master_service_run + 24 -> 11 indexer-worker 0x000000010d0b7001 main + 321 -> 12 libdyld.dylib 0x00007fff93d117e1 start + 0 -> 13 ??? 0x0000000000000001 0x0 + 1 12/4/12 12:34:17.286 PM dovecot[2660]: indexer: Error: Indexer worker disconnected, discarding 5 requests for dave 12/4/12 12:34:17.286 PM dovecot[2660]: indexer-worker(dave): Fatal: master: service(indexer-worker): child 2677 killed with signal 6 (core dumps disabled) 12/4/12 12:35:48.633 PM dovecot[2660]: indexer-worker(dave): Panic: file mailbox-list-fs.c: line 156 (fs_list_get_path): assertion failed: (mailbox_list_is_valid_pattern(_list, name)) 12/4/12 12:35:48.633 PM dovecot[2660]: indexer-worker(dave): Error: Raw backtrace: 2 libdovecot.0.dylib 0x0000000107b606c4 default_fatal_finish + 36 -> 3 libdovecot.0.dylib 0x0000000107b613e1 i_internal_fatal_handler + 49 -> 4 libdovecot.0.dylib 0x0000000107b6098e i_panic + 158 -> 5 libdovecot-storage.0.dylib 0x0000000107a95693 fs_list_get_path + 467 -> 6 indexer-worker 0x0000000107a3d341 master_connection_input + 513 -> 7 libdovecot.0.dylib 0x0000000107b6d5 12/4/12 12:35:48.634 PM dovecot[2660]: indexer-worker(dave): Error: ce io_loop_call_io + 46 -> 8 libdovecot.0.dylib 0x0000000107b6e737 io_loop_handler_run + 215 -> 9 libdovecot.0.dylib 0x0000000107b6d7e8 io_loop_run + 88 -> 10 libdovecot.0.dylib 0x0000000107b599c8 master_service_run + 24 -> 11 indexer-worker 0x0000000107a3d001 main + 321 -> 12 libdyld.dylib 0x00007fff93d117e1 start + 0 -> 13 ??? 0x0000000000000001 0x0 + 1 12/4/12 12:35:48.838 PM dovecot[2660]: indexer: Error: Indexer worker disconnected, discarding 5 requests for dave 12/4/12 12:35:48.838 PM dovecot[2660]: indexer-worker(dave): Fatal: master: service(indexer-worker): child 2694 killed with signal 6 (core dumps disabled) --8<---------------cut here---------------end--------------->8--- -- Dave Abrahams BoostPro Computing Software Development Training http://www.boostpro.com Clang/LLVM/EDG Compilers C++ Boost

2 1

[Dovecot] Dovecot version 1.2.17
by Philip Coackley 18 Dec '12

18 Dec '12

Hello, I am currently performing retrospective upgrades to an old ubuntu 9.10 Karmic Kola server. I need to upgrade dovecot 1.1.11 to 1.2.17 then to 2.0 etc. I have downloaded the binaries and run ./configure make make install. Dovecot compiles OK however I can not get the ./configure options to install at the following locations: executable: /usr/sbin/ config: /etc/dovecot/ Question: What are the ./configure options to do this? What are the ./configure options used to create the executable for unbuntu? (especially 9.10 if different). Reason for upgrade is auth-login errors every 2-3 weeks with 1.1.11 - looking at the release descriptions, this should fix this and the memory leaks! Phil Coackley

2 1

[Dovecot] Possible bug not sure on which end
by joshua.trier＠hostrocket.com 17 Dec '12

17 Dec '12

Hello Serveral of our clients are running into a problem relating to Gmails change in their "Check mail from other accounts (using POP3):". It seems related to the response dovecot sends back on a connection. Gmail is checking for a +OK in the header, however dovecot is sending OK without a prefix of +. Here is the response from Gmail webmail "Missing +OK response upon connecting to the server: * OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE AUTH=PLAIN] Dovecot ready." Here is dovecot header on a pop3-ssl connection * OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE AUTH=PLAIN AUTH=LOGIN] Dovecot ready. Who's end is this problem on? Google's, dovecots? Anyway to enable the response to include a +? Thanks Josh

3 2

Re: [Dovecot] Non-dovecot user looking for feedback
by Joseph Tam 17 Dec '12

17 Dec '12

On Mon, 17 Dec 2012, Chris Ross wrote: > I've been using UW-IMAPd for neigh on forever (at least since 2001). > But, as it's basically stalled, and I'm about to update the hardware > that is my mail server, I thought it was time to review existing > open-source IMAP servers. Dovecot is the top three in my looking. I converted a few years ago and never looked back. > So, am I right that dovecot supports only the "one big full file" mbox > format, and the maildir format? And if so, is it known to allow > multiple simultaneous IMAP clients to access and monitor and/or modify > the same folder simultaneously? I still run using mbox, and have not had a problem with corruption, however, this is probably owing to our modest IMAP demand and careful coordination of file locks among all the different MUAs. You're better off with Maildir and other formats, that are better designed for simultaneous access. There are other good reasons to switch formats. Dovecot also supports its own mdbox/sdbox for high performance: its a hybrid that tries to incorporate the best of both words. More info http://wiki2.dovecot.org/MailboxFormat > After those questions, it's just a "what do you think the pros and cons > of each are?" I can't speak to the other software you are looking at, but two factors I found to be particularly good: - flexibility: you can get almost every oddball setup to work. Migration from uw-imapd is fairly easy, but then again, uw-imapd didn't have a lot of knobs to turn. Plugins are available that extends its functionality. - support: this mailing list is great, and Timo is an absolute wizard in getting bugs fixed. (Note: I am shamelessly buttering him up so that he fixes the outstanding bug I reported!) Joseph Tam <jtam.home(a)gmail.com>

1 0

[Dovecot] more than one password_query
by goffer 16 Dec '12

16 Dec '12

Hello, I use dovecot + postfix with postfix sasl authentication by dovecot auth service. I have one question. Is it possible to have separate password_query only for auth service. I have two fields in database, one permits login to imap pop, second I want to use to permit login to smtp. Please explain how to set this (if it is possible) or give me some direction and sorry my English. -- View this message in context: http://dovecot.2317879.n4.nabble.com/more-than-one-password-query-tp39282.h… Sent from the Dovecot mailing list archive at Nabble.com.

1 1

[Dovecot] need help with dovecot-lda
by Andreas Meyer 16 Dec '12

16 Dec '12

Hello all! Some month ago I upgraded our dovecot installation from version 1.0.5 to version 2.1.7 without having any trouble. Postfix is delivering email directly per virtual transport to the maildirs and mailboxes of the users in /var/spool/vhosts/domains/.... Now I want to use dovecot-lda and created a new entry in the master.cf of postfix. dovecot unix - n n - - pipe flags=DRhu user=vmail:vmail argv=/usr/libexec/dovecot/dovecot-lda -f ${sender} -d ${recipient} It is not clear to me, how the dovecot-lda is started bei dovecot and you it finds the socket dovecot postfix created. The dovecot.conf looks like this; # 2.1.7: /etc/dovecot/dovecot.conf # OS: Linux 2.6.22.5-31-bigsmp i686 openSUSE 10.3 (i586) ext3 auth_verbose = yes disable_plaintext_auth = no log_path = /var/log/dovecot1 mail_location = mbox:/var/spool/vhosts/%d/dovecotprivate/%n:INBOX=/var/spool/vhosts/%d/%n !include /etc/dovecot/conf.d/*.conf passdb { args = /etc/dovecot/passwd driver = passwd-file } userdb { args = /etc/dovecot/passwd driver = passwd-file } ssl_cert = < /etc/dovecot/ssl/hostcert.pem ssl_key = < /etc/dovecot/ssl/hostkey.pem #verbose_ssl = yes plugin { quota = maildir:User quota quota_rule = *:storage=500MB # 10% of 1GB = 100MB quota_rule2 = Trash:storage=+10%% # 20% of 1GB = 200MB # quota_rule3 = Spam:storage=+20%% autocreate = Trash autocreate2 = Drafts autocreate3 = Sent autosubscribe = Trash autosubscribe2 = Drafts autosubscribe3 = Sent } I set soft_bounce = yes in main.cf of postfix and delivery of mail fails with Dec 14 16:13:34 delta postfix/virtual[14082]: BBC0F1B31294: to=<xyz(a)anup.de>, relay=virtual, delay=395, delays=395/0.01/0/0.13, dsn=4.1.1, status=SOFTBOUNCE (unknown user: "xyz(a)anup.de") This is the relevant part of the main.cf: dovecot_destination_recipient_limit = 1 virtual_mailbox_domains = anup.de bit-corner.de bitcorner.de baubetreuung-meyer.de saxelektro.de virtual_mailbox_base = /var/spool/vhosts #virtual_mailbox_maps = hash:/etc/postfix/vmailbox #virtual_minimum_uid = 100 #virtual_uid_maps = static:5000 #virtual_gid_maps = static:5000 virtual_mailbox_limit = 524288000 virtual_transport = dovecot I don't get it right. As far as I understand postfix uses virtual_transport = dovecot and hands the mail over to the socket dovecot. But how does dovecot know where to look for the socket. I am missing somthing but don't know what. Any help is appreciated! Andreas

3 27

[Dovecot] Dovecot "sleep" for short snapshots
by Peer Heinlein 15 Dec '12

15 Dec '12

Making consistent backups from Dovecot with mdbox requires a snapshoted filesystem with (at least) a short Downtime for Dovecot. I'd like to avoid killing several thenthousand IMAP-sessions for 60 seconds. Users don't like that... And the admin dislakes that too... What about a "short sleep" mechanism for Dovecot, where all writes are finished and new writes are easily delayed? I'd like to following process: *) "doveadm sleep" puts Dovecot in sleep-Mode: -- All existing writes are finished (before "doveadm sleep" is allowed to exit) -- All new writes are simply delayed -- All readonly-operations could work normally, so users can fetch and read mails *) The the admin can create a LVM-/SAN-snapshot of the filesystems *) "doveadm unsleep" will finish the magic Overall the whole process will take just some seconds. Users will recognize a simple write-delay, but they can work normally and the aren't unconnected at all. If the snapshot is really fast the users won't recognize the whole situation at all. Peer -- Heinlein Support GmbH Schwedter Str. 8/9b, 10119 Berlin http://www.heinlein-support.de Tel: 030 / 405051-42 Fax: 030 / 405051-19 Zwangsangaben lt. §35a GmbHG: HRB 93818 B / Amtsgericht Berlin-Charlottenburg, Geschäftsführer: Peer Heinlein -- Sitz: Berlin

2 1

[Dovecot] BUG: mishandling of username if it's a keyword?
by Jack Bates 14 Dec '12

14 Dec '12

Dec 14 14:33:03 test2 dovecot: auth: Debug: auth client connected (pid=24143) Dec 14 14:33:14 test2 dovecot: auth: Debug: client in: AUTH#0111#011PLAIN#011service=pop3#011secured#011session=D6dl6dDQdAAAAAAAAAAAAAAAAAAAAAAB#011lip=::1#011rip=::1#011lport=110#011rport=38004#011resp=<hidden> Dec 14 14:33:14 test2 dovecot: auth: Debug: static(home,::1,<D6dl6dDQdAAAAAAAAAAAAAAAAAAAAAAB>): lookup Dec 14 14:33:14 test2 dovecot: auth: Debug: client passdb out: OK#0111#011user=home Dec 14 14:33:14 test2 dovecot: auth: Debug: master in: REQUEST#0112033451009#01124143#0111#011243997dcca92c2dc1d1e401c78b4ea4f Dec 14 14:33:14 test2 dovecot: auth: Debug: master userdb out: USER#0112033451009#011uid=503#011gid=503#011home=/nfs/maildir/vmail/home#011mail_location=maildir:~/Maildir Dec 14 14:33:14 test2 dovecot: pop3-login: Login: user=<home>, method=PLAIN, rip=::1, lip=::1, mpid=24145, secured, session=<D6dl6dDQdAAAAAAAAAAAAAAAAAAAAAAB> Dec 14 14:33:14 test2 dovecot: pop3: Debug: Added userdb setting: mail_location=maildir:~/Maildir Dec 14 14:33:14 test2 dovecot: pop3(uid=503): Error: user uid=503: Couldn't drop privileges: User is missing UID (see mail_uid setting) Dec 14 14:33:14 test2 dovecot: pop3(uid=503): Error: Internal error occurred. Refer to server log for more information. compared to: Dec 14 14:37:25 test2 dovecot: pop3-login: Login: user=<home2>, method=PLAIN, rip=::1, lip=::1, mpid=24190, secured, session=<5Zdb+NDQdQAAAAAAAAAAAAAAAAAAAAAB> Dec 14 14:37:25 test2 dovecot: pop3: Debug: Added userdb setting: mail_location=maildir:~/Maildir Dec 14 14:37:25 test2 dovecot: pop3(home2): Debug: Effective uid=503, gid=503, home=/nfs/maildir/vmail/home2 userdb { args = uid=vmail gid=vmail home=/nfs/maildir/vmail/%u mail_location=maildir:~/Maildir driver = static } This was tested on a static passdb/userdb on a test server as well as production. version 2.1.10. When I have time, I'll dig into it myself after an upgrade to 2.1.12. Jack

2 5