- dovecot - dovecot.org

Dovecot 2.3.3 Mailbox does not exist
by Kunal A. 27 Feb '19

27 Feb '19

HI, I would deeply aprechiate if someone here could help me address a problem with ACL. I would personally refrain from e-mailing ever so often as I understand other users have more important issues to discuss about. Anyways I hope someone here could help. When I run ;- doveadm acl get -A Public/Archive I keep getting an error that says Mailbox does not exist . See erorr message below : - Error: Can't open mailbox Public/Archive: Mailbox doesn't exist: Public/Archive In the fastmail folder there is a folder called Archive, why isn't it being read? The folder is drwxr-xr-x. with vmail as owner. Deeply appreciate if someone could help with this. Many thanks dovecot -n output :- # 2.3.3 (dcead646b): /etc/dovecot/dovecot.conf # OS: Linux 4.20.10-200.fc29.x86_64 x86_64 Fedora release 29 (Twenty Nine) # Hostname: machine auth_mechanisms = plain login mail_location = maildir:~/Maildir mail_plugins = acl mail_privileged_group = mail mbox_write_locks = fcntl namespace { list = children location = maildir:/run/media/computer/Storage/Email/fastmail/ prefix = Public/ separator = / subscriptions = yes type = public } namespace inbox { inbox = yes list = 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 = INBOX/ separator = / } passdb { driver = pam } passdb { args = /etc/dovecot/dovecot-sql.conf.ext driver = sql } plugin { acl = vfile } postmaster_address = postmaster at example.com protocols = imap pop3 service auth-worker { user = vmail } service auth { unix_listener /var/spool/postfix/private/auth { group = postfix mode = 0666 user = postfix } unix_listener auth-userdb { mode = 0600 user = vmail } user = dovecot } service imap-login { inet_listener imap { port = 143 } inet_listener imaps { port = 993 ssl = yes } } service lmtp { unix_listener /var/spool/postfix/private/dovecot-lmtp { group = postfix mode = 0600 user = postfix } } service pop3-login { inet_listener pop3 { port = 0 } inet_listener pop3s { port = 995 ssl = yes } } ssl = required ssl_cert = </etc/pki/dovecot/certs/dovecot.pem ssl_cipher_list = PROFILE=SYSTEM ssl_key = # hidden, use -P to show it userdb { driver = passwd } userdb { args = uid=vmail gid=vmail home=/var/mail/vhosts/%d/%n driver = static } protocol imap { mail_plugins = acl imap_acl }

2 2

Re: performance issue with UID SEARCH
by Aki Tuomi 27 Feb '19

27 Feb '19

Without FTS, dovecot needs to open *each* and *every* email when doing text searches, which is understandably rather slow process. Aki On 27.2.2019 10.43, Marc Roos wrote: > > > I am not sure if this any help. From what I understand of maildir it has > lots of separate files, thus uid/gid lookups. Try running something like > nscd, that will cache these lookups? > > > > > -----Original Message----- > From: Aki Tuomi via dovecot [mailto:dovecot@dovecot.org] > Sent: 27 February 2019 06:24 > To: Ben Burke; Dovecot Mailing List > Subject: Re: performance issue with UID SEARCH > > > On 27 February 2019 03:27 Ben Burke via dovecot < > dovecot(a)dovecot.org> wrote: > > > Hi, > > I'm running dovecot 2.2.x and I'm having an issue where I see many > dovecot processes use all the available IO on a server. According > to > iotop the worst offenders seem to be in this state (NOTE: I swapped > in > phony username & IP info): > > dovecot/imap [someusername 123.456.789.012 UID SEARCH] > > The server in question is running with Maildirs on top of an XFS > filesystem. Is there anything I can do to optimize "UID SEARCH" or > find > out why it's being a problem? I've read > https://wiki2.dovecot.org/PerformanceTuning and the linked pages. > > By "being a problem" I mean iostat -xmt 1 /dev/diskdevice shows > 100% > utilization for long periods and in some cases io service times are > > taking many seconds... which causes thunderbird to timeout when > doing > things like appending messages to user "Sent" mailboxes. > > Any ideas? > > Thanks, > Ben Burke > > > Are you using FTS? If not, you should. See > https://wiki.dovecot.org/Plugins/FTS > --- > Aki Tuomi > >

1 0

CASifying dovecot and webmail
by Juan Carlos Sánchez 27 Feb '19

27 Feb '19

Hello: I'm trying to authenticate webmail (roundcube) and dovecot with Apereo CAS. I don't find any good guide and everything I have found is some years old. Has any of you done a recent integration with CAS and can share the details ? Best regards -- ------------------------------------------------------ Juan Carlos Sanchez Hernandez Servicio de Infraestructura e Innovación Universidad Politecnica de Madrid - www.upm.es Rectorado Avda. Ramiro de Maeztu 7 28040 Madrid ------------------------------------------------------

3 2

Re: Linking mailboxes
by Aki Tuomi 27 Feb '19

27 Feb '19

drwx-----x kunal kunal Storage I'm guessing this point is being more restrictive. Can you try changing this into 0755? Also suddenly you are no longer using vmail but random user IDs. This only works with vmail. Aki > On 25 February 2019 19:56 Kunal A. <kunal.oa(a)gmail.com> wrote: > > > Hi, > Yes. Output ;- > > dr-xr-xr-x root root / > drwxr-xr-x root root run > drwxr-xr-x root root media > drwxr-x--x root root computer > drwx-----x kunal kunal Storage > drwx------ vmail vmail Email > drwx------ vmail vmail tmp > > > > On Mon, Feb 25, 2019 at 12:53 PM Aki Tuomi <aki.tuomi(a)open-xchange.com> wrote: > > Can you run > > > > namei -vl /run/media/computer/Storage/Email/tmp > > > > the + at the end does not mean that, it means there is an extended POSIX ACL. > > > > Aki > > > > > On 25 February 2019 19:50 Kunal A. <kunal.oa(a)gmail.com> wrote: > > > > > > > > > Hey Aki, > > > Sorry for the trouble but I am still unable to get it working. The file permissions for upper files are now drwxr-xr-x. but I think they need to be drwxr-x--x+ ? > > > Any idea how to change it to that? > > > > > > > > > Error log is below:- > > > > > > doveadm(lightdm): Error: Can't open mailbox Storage/Email: stat(/run/media/computer/Storage/Email/tmp) failed: Permission denied (euid=976(lightdm) egid=976(lightdm) missing +x perm: /run/media/computer/Storage/Email, dir owned by 5000:5000 mode=0700) > > > doveadm(setroubleshoot): Error: stat(/run/media/computer/Storage/Email/tmp) failed: Permission denied (euid=975(setroubleshoot) egid=975(setroubleshoot) missing +x perm: /run/media/computer/Storage/Email, dir owned by 5000:5000 mode=0700) > > > doveadm(setroubleshoot): Error: Can't open mailbox Storage/Email: stat(/run/media/computer/Storage/Email/tmp) failed: Permission denied (euid=975(setroubleshoot) egid=975(setroubleshoot) missing +x perm: /run/media/computer/Storage/Email, dir owned by 5000:5000 mode=0700) > > > doveadm(dovenull): Error: stat(/run/media/computer/Storage/Email/tmp) failed: Permission denied (euid=974(dovenull) egid=973(dovenull) missing +x perm: /run/media/computer/Storage/Email, dir owned by 5000:5000 mode=0700) > > > ~ > > > > > > > > > > > > > > > On Mon, Feb 25, 2019 at 11:58 AM Kunal A. <kunal.oa(a)gmail.com> wrote: > > > > Dear Aki, > > > > Many thanks again. Thanks for explaining this. > > > > Very grateful for your assistance. > > > > Thanks a bunch ! > > > > Kunal > > > > > > > > > > > > > > > > On Mon, Feb 25, 2019 at 11:56 AM Aki Tuomi <aki.tuomi(a)open-xchange.com> wrote: > > > > > At minimum directories > > > > > > > > > > /run/media/computer/Storage/ > > > > > > > > > > need to have o+x, so you can do > > > > > > > > > > chmod o+x /run /run/media /run/media/computer /run/media/computer/Storage > > > > > > > > > > and > > > > > > > > > > chown -R vmail:vmail /run/media/computer/Storage/Email > > > > > > > > > > Aki > > > > > > > > > > > On 25 February 2019 18:47 Kunal A. via dovecot <dovecot(a)dovecot.org> wrote: > > > > > > > > > > > > > > > > > > Dear Aki, > > > > > > I have given permission to the folders Storage/Email to vmail. > > > > > > But the upper hierarchy is not under vmail . > > > > > > > > > > > > Full directory = /run/media/computer/Storage/Email > > > > > > > > > > > > /run/media/computer = not vmail > > > > > > Storage/Email = vmail > > > > > > > > > > > > How do I get this corrected ? It seems like vmail needs access through all through run/media/computer . > > > > > > > > > > > > Please excuse my ignorance as I am not very familiar with filesystem permissions. :-/ > > > > > > > > > > > > Sincere apologies for the trouble. > > > > > > > > > > > > Thanks > > > > > > Kunal A. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, Feb 25, 2019 at 11:36 AM Aki Tuomi <aki.tuomi(a)open-xchange.com> wrote: > > > > > > > I would only run it for the mail folders to be safe. > > > > > > > > > > > > > > Aki > > > > > > > > > > > > > > > On 25 February 2019 18:34 Kunal A. via dovecot <dovecot(a)dovecot.org> wrote: > > > > > > > > > > > > > > > > > > > > > > > > Dear Aki, > > > > > > > > Many thanks for your feedback. Its what I figured I should do but I was not sure how safe it was. > > > > > > > > > > > > > > > > Would it make sense to execute the following to /run/media/computer so that the user/group is vmail? > > > > > > > > sudo chown -R vmail:vmail to the /run folder ? > > > > > > > > > > > > > > > > Also if this practice is safe to perform? > > > > > > > > > > > > > > > > Thanks > > > > > > > > Kunal A > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, Feb 25, 2019 at 11:20 AM Aki Tuomi <aki.tuomi(a)open-xchange.com> wrote: > > > > > > > > > > > > > > > > > > > On 25 February 2019 18:19 Aki Tuomi via dovecot <dovecot(a)dovecot.org> wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On 25 February 2019 17:56 Kunal A. via dovecot <dovecot(a)dovecot.org> wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Hey Guys! > > > > > > > > > > > Many thanks for your feedback. I have kind of gotten it to work but I just struggling to get the permissions of the folder correct. Highly appreciate if someone could help me set the correct permission settings for it . Terribly sorry for my ignorance. > > > > > > > > > > > > > > > > > > > > > > My error msg is listed below : - > > > > > > > > > > > > > > > > > > > > > > doveadm(vmail): Error: stat(/run/media/computer/Storage/Email/tmp) failed: Permission denied (euid=5000(vmail) egid=5000(vmail) missing +x perm: /run/media/computer, dir owned by 0:0 mode=0750) > > > > > > > > > > > doveadm(vmail): Error: Can't open mailbox Storage/Email: stat(/run/media/computer/Storage/Email/tmp) failed: Permission denied (euid=5000(vmail) egid=5000(vmail) missing +x perm: /run/media/computer, dir owned by 0:0 > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > You need to give write permissions to vmail on /run/media/computer/Storage/Email > > > > > > > > > > > > > > > > > > > > You can > > > > > > > > > > - change owner to vmail > > > > > > > > > > - change group to vmail and add rwx for group > > > > > > > > > > - use posix ACLs > > > > > > > > > > - change mode to 0777 (unsafe) > > > > > > > > > > > > > > > > > > > > Aki > > > > > > > > > > > > > > > > > > And the same applies for all files & directories under /run/media/computer/Storage/Email > > > > > > > > > > > > > > > > > > Aki > > > > > > > > > > > > > > > > > > > > > > >

3 2

performance issue with UID SEARCH
by Ben Burke 27 Feb '19

27 Feb '19

Hi, I'm running dovecot 2.2.x and I'm having an issue where I see many dovecot processes use all the available IO on a server. According to iotop the worst offenders seem to be in this state (NOTE: I swapped in phony username & IP info): dovecot/imap [someusername 123.456.789.012 UID SEARCH] The server in question is running with Maildirs on top of an XFS filesystem. Is there anything I can do to optimize "UID SEARCH" or find out why it's being a problem? I've read https://wiki2.dovecot.org/PerformanceTuning and the linked pages. By "being a problem" I mean iostat -xmt 1 /dev/diskdevice shows 100% utilization for long periods and in some cases io service times are taking many seconds... which causes thunderbird to timeout when doing things like appending messages to user "Sent" mailboxes. Any ideas? Thanks, Ben Burke

2 1

new Centos server install yum dependancy error
by Voytek Eymont 26 Feb '19

26 Feb '19

I'm installing a new Centos 7 server, following repo.dovecot.org/ page, but, get this: (I've opted to install all BUT pgsql (use mysql), not sure I really need all other packages..?) should I just go for "dovecot dovecot-mysql dovecot-pigeonhole" ?? tried # yum install dovecot dovecot-debuginfo dovecot-devel dovecot-imaptest dovecot-imaptest-debuginfo dovecot-lua dovecot-mysql dovecot-pigeonhole dovecot-pigeonhole-debuginfo dovecot-pigeonhole-devel ... --> Finished Dependency Resolution Error: Package: 2:dovecot-lua-2.3.4.1-1.x86_64 (dovecot-2.3-latest) Requires: dovecot = 2:2.3.4.1-1 Available: 1:dovecot-2.2.36-3.el7.i686 (base) dovecot = 1:2.2.36-3.el7 Error: Package: 2:dovecot-pigeonhole-devel-2.3.4.1-1.x86_64 (dovecot-2.3-latest) Requires: dovecot-pigeonhole = 2:2.3.4.1-1 Installing: 1:dovecot-pigeonhole-2.2.36-3.el7.x86_64 (base) dovecot-pigeonhole = 1:2.2.36-3.el7 Error: Package: 2:dovecot-imaptest-2.3.4.1-1.x86_64 (dovecot-2.3-latest) Requires: dovecot = 2:2.3.4.1-1 Available: 1:dovecot-2.2.36-3.el7.i686 (base) dovecot = 1:2.2.36-3.el7 You could try using --skip-broken to work around the problem You could try running: rpm -Va --nofiles --nodigest -- Voytek

4 12

Removing a mailbox from a dovecot cluster
by Francis 26 Feb '19

26 Feb '19

Hello, How should I remove a mailbox from a dovecot cluster? I have two dovecot in a cluster. I tried to remove one user's mailbox with theses commands I found with a little google search: doveadm flags add -u user(a)domain.tld "\\Deleted" ALL doveadm expunge -u user(a)domain.tld mailbox "*" DELETED doveadm purge -u user(a)domain.tld rm -fr /var/vmail/domains/domain.tld/user If I do theses commands in a script (so one command after one, very quick), the data just keep coming back from the other server in the cluster (I suppose this is what's happening). If I do theses commands manually, so more slowly, the mailbox is created again but empty this time. If I connect from an IMAP client, I see an empty mailbox. Same thing if I connect an IMAP client to the second server from the cluster. But I still see the data in the storage (mdbox) of the second server (the "m.#" files are still there an they waste space on disks). I tried to do a "doveadm purge" on the second server, but it did nothing more, the data is still keep on the disks. I'm running dovecot 2.2.33.2 on ubuntu 18.04 on both servers. Thanks! -- Francis

1 1

Virus scan + removal on a mdbox mail storage
by Christoph Haas 26 Feb '19

26 Feb '19

Hi, I need advice on how virus scan and removal can be done on a _mdbox_ mail storage? On a maildir storage the virus scanner (e.g. clamav etc.) can detect and remove a email that is infected, since every email and attachment are stored in separate files. But in mdbox the emails and attachments are compressed together in one ore more mdbox-files ... I am anxious to convert my mail storage for virus scanning into maildir format, since I don't know if a virus or crypto trojan con be activated with this converting action =:-o Cheers Christoph. -- Christoph Haas

3 6

Linking mailboxes
by Kunal A. 25 Feb '19

25 Feb '19

Hi, I am new to Dovecot and I am trying to link an email account folder to another Maildir(its a backup of the older email). I want the email account to be able to access the older emails. I am assuming this is done by creating a symlink? Currently reading instructions on this page:- https://wiki.dovecot.org/SharedMailboxes/Symlinks At the moment I have the folder:- /var/mail/vhosts/example.com/email1/Maildir/.INBOX.newfolder Inside the folder:- drwx------. 2 vmail vmail 4096 Feb 23 13:39 cur -rw-------. 1 vmail vmail 232 Feb 23 13:43 dovecot.index.log -rw-------. 1 vmail vmail 51 Feb 23 13:43 dovecot-uidlist -rw-------. 1 vmail vmail 0 Feb 23 13:39 maildirfolder drwx------. 2 vmail vmail 4096 Feb 23 13:39 new lrwxrwxrwx. 1 root root 45 Feb 23 13:46 Other -> /run/media/computer/Storage/Email/mail/Other drwx------. 2 vmail vmail 4096 Feb 23 13:39 tmp lrwxrwxrwx. 1 root root 5 Feb 23 13:46 Users -> Users I am not able to access the folders in the Other folder structure(inside other there are other email inboxes ie(email1@example.com,email2@example.com,etc.). I want to get access to all the email subfolders. Could someone here advise how to do this or point me to some tutorials for this. Sincere apologies for this and many thanks in advance for helping... Dovecot -n output :- # 2.3.3 (dcead646b): /etc/dovecot/dovecot.conf # OS: Linux 4.20.10-200.fc29.x86_64 x86_64 Fedora release 29 (Twenty Nine) # Hostname: computer auth_mechanisms = plain login mail_location = maildir:~/Maildir mail_privileged_group = mail mbox_write_locks = fcntl 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 } passdb { args = /etc/dovecot/dovecot-sql.conf.ext driver = sql } postmaster_address = postmaster at example.com protocols = imap pop3 service auth-worker { user = vmail } service auth { unix_listener /var/spool/postfix/private/auth { group = postfix mode = 0666 user = postfix } unix_listener auth-userdb { mode = 0600 user = vmail } user = dovecot } service imap-login { inet_listener imap { port = 143 } inet_listener imaps { port = 993 ssl = yes } } service lmtp { unix_listener /var/spool/postfix/private/dovecot-lmtp { group = postfix mode = 0600 user = postfix } } service pop3-login { inet_listener pop3 { port = 0 } inet_listener pop3s { port = 995 ssl = yes } } ssl = required ssl_cert = </etc/pki/dovecot/certs/dovecot.pem ssl_cipher_list = PROFILE=SYSTEM ssl_key = # hidden, use -P to show it userdb { driver = passwd } userdb { args = uid=vmail gid=vmail home=/var/mail/vhosts/%d/%n driver = static }

3 12

Corrupted transaction log and redownloading
by azurit＠pobox.sk 25 Feb '19

25 Feb '19

Hi, today we had to hard-reboot one of our dovecot proxy backends and after boot up, lot's of accounts were showing this error and appeard empty: Error: Corrupted transaction log file ... After deleting the indexes (not uid file, problem wasn't in deleting indexes, i'm doing this often), e-mails showed up ok but e-mail clinets started to download them again. Why this happened and how to prevent it (other than 'not hard-rebooting')? I know that reboot could cause corruption of uid files in some cases but this time it looks like this was caused by corrupted indexes - how index corruption can cause redownloading and why? Thanks for any info. azur

2 1