Re: cyrus sieve redirect action and SPF
On Thu, 17 Apr 2014 12:34:37 +0100 Geoff Winkless wrote: > On 17 April 2014 12:26, Frank Elsner wrote: > > > My cyrus-imapd-2.3.16-6.el6_2.5.x86_64 uses . > > > > I'd like to use the original envelope sender even if it doesn't play with > > SPF. > > > The MTA won't let you set envelope-sender because it's a security hole, so > even though cyrus is saying "pretend that this email is from > x...@somedomain.com" the MTA sets it as cyrus@FQDN instead. > > If you're using exim, add cyrus to your trusted_users line. That worked > for me. Oh shit, why did this not came to my mind? Thanks, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
Re: cyrus sieve redirect action and SPF
On Thu, 17 Apr 2014 12:27:37 +0300 Deniss wrote: > Hello, Hello, > Original envelope From of an email is used when redirect is set in sieve > to resend the mail to another email address. really? My cyrus-imapd-2.3.16-6.el6_2.5.x86_64 uses . I'd like to use the original envelope sender even if it doesn't play with SPF. Any hint? Kind regards, Frank Elsner (TU Berlin Postmaster) Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
cyr_expire (2.3.16) problem
Hello, from time to time we have this type of problem during the nightly cyr_expire: Feb 10 04:01:50 mailbackend-4 cyrus-backend/cyr_expire[15569]: IOERROR: user. zero index/expunge record 299/301 Feb 10 04:01:50 mailbackend-4 cyrus-backend/cyr_expire[15569]: failure expiring user.: System I/O error Feb 11 04:01:48 mailbackend-4 cyrus-backend/cyr_expire[2010]: IOERROR: user. zero index/expunge record 299/301 Feb 11 04:01:48 mailbackend-4 cyrus-backend/cyr_expire[2010]: failure expiring user.: System I/O error Feb 12 04:02:19 mailbackend-4 cyrus-backend/cyr_expire[27580]: IOERROR: user. zero index/expunge record 299/301 Feb 12 04:02:19 mailbackend-4 cyrus-backend/cyr_expire[27580]: failure expiring user.: System I/O error Feb 13 04:03:15 mailbackend-4 cyrus-backend/cyr_expire[19753]: IOERROR: user. zero index/expunge record 299/301 Feb 13 04:03:15 mailbackend-4 cyrus-backend/cyr_expire[19753]: failure expiring user.: System I/O error This leaves cyrus.*.NEW files in the mailbox directory of the user and requires a reconstruct. What's behind? And how can we avoid this problem? --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
Re: How can this happen?
On Fri, 25 Jan 2013 10:48:28 +0100 Simon Matter wrote: > > On Thu, 24 Jan 2013 13:12:33 -0800 (PST) Andrew Morgan wrote: > >> On Thu, 24 Jan 2013, Frank Elsner wrote: > >> > >> > > >> > Hello, > >> > > >> > we have the strange situation with our murder environment that > >> > > >> > mailbackend has > >> > > >> > user.x.Sent2 default x lrswipkxtecda > >> > > >> > but the folder is non-existent in the filesystem. > >> > > >> > The mupdate server doesn't know this folder (not in mailboxes.db) > >> > > >> > How can this happen? > >> > >> Some bug in Cyrus? > > > > Don't know. > > > >> If you want to fix this, you can try creating the proper structure on > >> the > >> filesystem, run reconstruct to get Cyrus to sync back up with it, then > >> delete the folder using cyradm or an IMAP client. > > > > I didn't ask for a fix. I ask for a possible reason. > > I suggest to at lest report the version of Cyrus used. Oh, shit. Shame on me. 2.3.16 on RHEL. --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
Re: How can this happen?
On Thu, 24 Jan 2013 13:12:33 -0800 (PST) Andrew Morgan wrote: > On Thu, 24 Jan 2013, Frank Elsner wrote: > > > > > Hello, > > > > we have the strange situation with our murder environment that > > > > mailbackend has > > > > user.x.Sent2 default x lrswipkxtecda > > > > but the folder is non-existent in the filesystem. > > > > The mupdate server doesn't know this folder (not in mailboxes.db) > > > > How can this happen? > > Some bug in Cyrus? Don't know. > If you want to fix this, you can try creating the proper structure on the > filesystem, run reconstruct to get Cyrus to sync back up with it, then > delete the folder using cyradm or an IMAP client. I didn't ask for a fix. I ask for a possible reason. We cleaned the situation by removing the entry from the mailboxes.db on the mailbackend using cyr_dbtool. --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
How can this happen?
Hello, we have the strange situation with our murder environment that mailbackend has user.x.Sent2 default x lrswipkxtecda but the folder is non-existent in the filesystem. The mupdate server doesn't know this folder (not in mailboxes.db) How can this happen? Greetings, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
Re: successful create but unsuccessful subscribe
On Wed, 19 Dec 2012 13:45:36 -0800 (PST) Andrew Morgan wrote: > On Wed, 19 Dec 2012, Frank Elsner wrote: > > > On Wed, 19 Dec 2012 10:10:43 -0800 (PST) Andrew Morgan wrote: > > > > [ ... ] > > > >> I don't know why, but we have always operated with prefork=1 here. As far > >> as I can tell, it never runs more than 1 mupdate process. That single > >> mupdate process has 15 connections in total from our 3 frontend servers. > >> There doesn't seem to be a need for it to spawn additional mupdate > >> processes. Now that I look closer, I see that mupdate is threaded... > >> > >> We have 3 frontends and 3 backends. Each backend has about 20,000 users > >> on it. > >> > >> Here is my cyrus.conf entry on the mupdate master: > >> > >>mupdate cmd="/usr/local/cyrus/bin/mupdate -m" listen="3905" > >> proto="tcp4" prefork=1 > >> > >> and on the frontends: > >> > >>mupdate cmd="/usr/local/cyrus/bin/mupdate" listen="3905" > >> proto="tcp4" prefork=1 > >> [ ... ] > > Ok, sounds good. On the mupdate master we already have the mupdate_* > > settings. > > Shall we put mupdate_* settings into the imapd.conf on the frontends too? > > I have never set those on either mupdate master or frontends, although I > have the defaults in imapd.conf commented out. At least here with our > systems, the defaults seem to be working well. We have tested with "mupdate ... prefork=1" on master and frontend(s) and it started ok, but we are facing an other problem: On our RHEL system we have encreased the limit on processes and open files for the user "cyrus" in /etc/security/limits.de/cyrus-imapd.conf by the lines cyrussoftnproc 3072 cyrussoftnofile 8192 cyrushardnofile 8192 and we put a "ulimit -u 3072 -n 8192" right before the start of cyrus-master in /etc/init.d/cyrus-imapd. The running cyrus-master has these limits but the mupdate process only has 256 open files allowed. This leads to Dec 20 11:44:34 mupdate-2 mupdate[13465]: accept failed: Too many open files Dec 20 11:52:19 mupdate-2 mupdate[14098]: socket: Too many open files Any pointer towards a solution is welcome. Just for the records: The mupdate master has this legacy settings mupdate_connections_max: 256 mupdate_workers_max: 512 mupdate_workers_start: 75 mupdate_workers_minspare: 75 mupdate_workers_maxspare: 120 mupdate_config: standard mupdate_port: 3905 --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
Re: successful create but unsuccessful subscribe
On Wed, 19 Dec 2012 10:10:43 -0800 (PST) Andrew Morgan wrote: [ ... ] > I don't know why, but we have always operated with prefork=1 here. As far > as I can tell, it never runs more than 1 mupdate process. That single > mupdate process has 15 connections in total from our 3 frontend servers. > There doesn't seem to be a need for it to spawn additional mupdate > processes. Now that I look closer, I see that mupdate is threaded... > > We have 3 frontends and 3 backends. Each backend has about 20,000 users > on it. > > Here is my cyrus.conf entry on the mupdate master: > >mupdate cmd="/usr/local/cyrus/bin/mupdate -m" listen="3905" > proto="tcp4" prefork=1 > > and on the frontends: > >mupdate cmd="/usr/local/cyrus/bin/mupdate" listen="3905" > proto="tcp4" prefork=1 > > Hmm, there is no manpage for mupdate either! Digging around in the source > code shows that there are configuration options for mupdate in imapd.conf, > such as: > > mupdate_workers_max: 50 > The maximum number of mupdate worker threads (overall) > > mupdate_workers_maxspare: 10 > The maximum number of idle mupdate worker threads > > mupdate_workers_minspare: 2 > The minimum number of idle mupdate worker threads > > mupdate_workers_start: 5 > The number of mupdate worker threads to start Ok, sounds good. On the mupdate master we already have the mupdate_* settings. Shall we put mupdate_* settings into the imapd.conf on the frontends too? --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
Re(2): successful create but unsuccessful subscribe
On Mon, 17 Dec 2012 11:11:13 +0100 Frank Elsner wrote: [ ... ] > On startup all frontends get the complete mailboxlist from the mupdate server. This is the log from one of the mupdate processes after "service cyrus-imapd restart": Dec 17 11:27:53 mailfrontend-5 cyrus/mupdate[22260]: executed Dec 17 11:27:53 mailfrontend-5 cyrus/mupdate[22260]: successful mupdate connection to mupdate-2-febe.intern.tu-berlin.de Dec 17 11:27:53 mailfrontend-5 cyrus/mupdate[22260]: scarfing mailbox list from master mupdate server Dec 17 11:27:57 mailfrontend-5 cyrus/mupdate[22260]: unready for connections Dec 17 11:27:57 mailfrontend-5 cyrus/mupdate[22260]: synchronizing mailbox list with master mupdate server Dec 17 11:28:00 mailfrontend-5 cyrus/mupdate[22260]: mailbox list synchronization complete Dec 17 11:28:01 mailfrontend-5 cyrus/mupdate[22260]: New worker thread started, for a total of 1 Dec 17 11:28:01 mailfrontend-5 cyrus/mupdate[22260]: New worker thread started, for a total of 2 Dec 17 11:28:01 mailfrontend-5 cyrus/mupdate[22260]: New worker thread started, for a total of 3 Dec 17 11:28:01 mailfrontend-5 cyrus/mupdate[22260]: New worker thread started, for a total of 4 Dec 17 11:28:01 mailfrontend-5 cyrus/mupdate[22260]: New worker thread started, for a total of 5 Dec 17 11:29:01 mailfrontend-5 cyrus/mupdate[22260]: Thread timed out waiting for listener_lock Dec 17 11:29:01 mailfrontend-5 cyrus/mupdate[22260]: Worker thread finished, for a total of 4 (4 spare) Dec 17 11:29:01 mailfrontend-5 cyrus/mupdate[22260]: Thread timed out waiting for listener_lock Dec 17 11:29:01 mailfrontend-5 cyrus/mupdate[22260]: Worker thread finished, for a total of 3 (3 spare) Dec 17 11:29:01 mailfrontend-5 cyrus/mupdate[22260]: Thread timed out waiting for listener_lock Dec 17 11:29:01 mailfrontend-5 cyrus/mupdate[22260]: Worker thread finished, for a total of 2 (2 spare) What does "unready for connections" mean? "a total of 2" for test only. standard is prefork=5. --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
Re: successful create but unsuccessful subscribe
On Fri, 14 Dec 2012 09:50:50 +0100 Kerstin Espey wrote: > On 13.12.2012 18:13, Dan White wrote: > > On 12/13/12 11:50 +0100, Kerstin Espey wrote: > >> we see a strange behaviour in our dev cyrus murder environment which we > >> can't explain. > >> > >> We run cyrus 2.4 on ubuntu server. > > > > What specific release are you using? > > 2.4.12-2 on ubuntu precise. As already mentioned, we - TU-Berlin.DE - have the same problem. We run a mupdate-server (2.3.16, RHEL), 3 Frontends (2.3.16, Solaris 10), 4 Frontends (2.3.16, RHEL) to replace the Solaris hosts in the future. 4 Backends (2 Solaris 10, 2 RHEL) All frontends run the same configuration. On startup all frontends get the complete mailboxlist from the mupdate server. After a while the RHEL hosts show the same behavior as Kerstins hosts. I case the mailbox list on a (RHEL-)frontend is not identical to the one on the mupdate server, a "kill mupdate" forces a complete an the frontend which is successful. We have no problem on the communication from the backends to the mupdate server. Folders created on one of the backends immediatly show up on the mupdate server. Just my 0.02 €. Any pointer to find the problem or even solve is welcome. --Frank Elsner (TU Berlin Postmaster) Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
Re: successful create but unsuccessful subscribe
On Thu, 13 Dec 2012 11:13:21 -0600 Dan White wrote: [ ... ] > Did this work for you differently in a previous version of cyrus? We have exactly the same problem with cyrus 2.3.16 on RHEL. --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
2nd try: xfermailbox oddities
I've asked this already on Fri, 7 Sep 2012 but didn't a respose until now. So this is another try: Hello, we need to migrate mailboxes from 2 backend hosts, Solaris 10, cyrus-2.3.16 to 2 backend hosts, RedHat Enterprise Linux, cyrus-2.3.16 in an murder environment. During our initial testing we discovered the following problem: "xfermailmailbox" transfers the user's mailbox and sieve filter just fine from source (Solaris) to destination (RHEL). But all mails in the "xfer"ed mailbox show up as unread on the destination host, no matter which client used. Futhermore the seen file on the source host remains there, the seen file on the destination host is somewhat smaller. What's going on? - Why show the mails up as unread? - Why isn't the seen file "xfer"ed as is? Thanks in advance, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
xfermailbox oddities
Hello, we need to migrate mailboxes from 2 backend hosts, Solaris 10, cyrus-2.3.16 to 2 backend hosts, RedHat Enterprise Linux, cyrus-2.3.16 in an murder environment. During our initial testing we discovered the following problem: "xfermailmailbox" transfers the user's mailbox and sieve filter just fine from source (Solaris) to destination (RHEL). But all mails in the "xfer"ed mailbox show up as unread on the destination host, no matter which client used. Futhermore the seen file on the source host remains there, the seen file on the destination host is somewhat smaller. What's going on? - Why show the mails up as unread? - Why isn't the seen file "xfer"ed as is? Thanks in advance, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ To Unsubscribe: https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus
Re: Lock problem with mupdate
On Tue, 17 Apr 2012 13:29:41 +0200 Simon Matter wrote: > > On Tue, 17 Apr 2012 05:43:41 -0400 Dave McMurtrie wrote: > >> strace the server and find out what it's doing. Look at the logs on the > >> server. Run netstat to confirm both sides still have an established > >> connection and that iptables isn't silently dropping packets, etc. > > > > No iptables active, due to private net. > > SELinux? Same answer as above I guess? [root@eg-mailfrontend elsnccpa]# getenforce Disabled --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Re: Lock problem with mupdate
On Tue, 17 Apr 2012 05:43:41 -0400 Dave McMurtrie wrote:
> strace the server and find out what it's doing. Look at the logs on the
> server. Run netstat to confirm both sides still have an established
> connection and that iptables isn't silently dropping packets, etc.
No iptables active, due to private net.
On the mupdate server:
mupdate-febe.intern.tu-berlin.de.mupdate
eg-mailfrontend-febe.intern.elgouna.tu-berlin.de.40536 14656 0 49248
0 ESTABLISHED
mupdate-febe.intern.tu-berlin.de.mupdate
eg-mailfrontend-febe.intern.elgouna.tu-berlin.de.40537 14656 0 49248
0 ESTABLISHED
mupdate-febe.intern.tu-berlin.de.mupdate
eg-mailfrontend-febe.intern.elgouna.tu-berlin.de.40538 69568 0 49248
0 ESTABLISHED
mupdate-febe.intern.tu-berlin.de.mupdate
eg-mailfrontend-febe.intern.elgouna.tu-berlin.de.40539 64128 0 49248
0 ESTABLISHED
On eg-mailfrontend:
[root@eg-mailfrontend elsnccpa]# netstat | grep mupdate
tcp0 0 eg-mailfrontend-febe.:40538 172.26.17.69:mupdate
ESTABLISHED
tcp0 0 eg-mailfrontend-febe.:40536 172.26.17.69:mupdate
ESTABLISHED
tcp0 0 eg-mailfrontend-febe.:40539 172.26.17.69:mupdate
ESTABLISHED
tcp0 0 eg-mailfrontend-febe.:40537 172.26.17.69:mupdate
ESTABLISHED
Furthermore: If done "nanually" it seems to work:
[root@eg-mailfrontend elsnccpa]# mupdatetest -a murder
mupdate-febe.intern.tu-berlin.de
S: * AUTH "CRAM-MD5"
S: * PARTIAL-UPDATE
S: * OK MUPDATE "mupdate" "Cyrus Murder" "v2.3.11" "(master)"
C: A01 AUTHENTICATE "CRAM-MD5"
S: PDE1MDY2MjE4NDkuOTI2MDg1OEBtdXBkYXRlPg==
Please enter your password:
C: bXVyZGVyIDA4OTNkYzVjNWU0NTI0OTAyZTYyMTM4NDI0OWJiNWVI
S: A01 OK "Authenticated"
Authenticated.
Security strength factor: 0
. update
. MAILBOX {13+}
user.a-1991-l {46+}
mailbackend-2-febe.intern.tu-berlin.de!default {23+}
a-1991-llrswipkxtecda
. MAILBOX {12+}
user.a-ahmed {46+}
mailbackend-1-febe.intern.tu-berlin.de!default {22+}
a-ahmed lrswipkxtecda
. MAILBOX {25+}
user.a-ahmed.Entw&APw-rfe {46+}
mailbackend-1-febe.intern.tu-berlin.de!default {22+}
.. and so on ...
--Frank Elsner
Cyrus Home Page: http://www.cyrusimap.org/
List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Re: Lock problem with mupdate
On Tue, 17 Apr 2012 04:41:57 -0400 Dave McMurtrie wrote: > lsof so you can find out what file descriptor 9 is that read() is blocking > on. If it's the socket to your mupdate server, figure out why it isn't > responding. Thanks. Indeed it is the connection to the mupdate server: mupdate 7395 cyrus9u IPv43273817 0t0 TCP eg-mailfrontend-febe.intern.elgouna.tu-berlin.de:40362->172.26.17.69:mupdate (ESTABLISHED) So we have to investigate why we can ping in both directions, why the frontend can autheticate on the mupdate server and why the answer doesn't get through. Thanks so far, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Re: Lock problem with mupdate
On Tue, 17 Apr 2012 03:17:12 -0400 Dave McMurtrie wrote: > mupdate is multithreaded. Try strace -f -p to see what it's doing. Process 7393 attached with 2 threads - interrupt to quit [pid 7398] read(9, [pid 7393] futex(0x7fdc39d53e84, FUTEX_WAIT_PRIVATE, 1, NULL Nearly the same info, not very informative :-( --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Lock problem with mupdate
Hi experts, we have a mupdate server (Solaris 10, Cyrus 2.3.16) and frontnend and backend servers (Redhat Linux due to migration from Solaris environment). The RHEL frontend connects to the (Solaris) mupdate server, authentication work. But after Apr 17 07:27:21 eg-mailfrontend cyrus/mupdate[7395]: successful mupdate connection to mupdate-febe.intern.tu-berlin.de Apr 17 07:27:21 eg-mailfrontend cyrus/mupdate[7395]: scarfing mailbox list from master mupdate server nothing happens, the mailbox list lacks update. Lets look what mupdate does: [root@eg-mailfrontend elsnccpa]# ps -ef | grep mupdate | grep -v grep cyrus 7393 7388 0 07:27 ?00:00:00 mupdate cyrus 7395 7388 0 07:27 ?00:00:00 mupdate [root@eg-mailfrontend elsnccpa]# strace -p 7396 Process 7396 attached - interrupt to quit accept(4, ^C Process 7396 detached [root@eg-mailfrontend elsnccpa]# man tcpdump [root@eg-mailfrontend elsnccpa]# man imapd.conf [root@eg-mailfrontend elsnccpa]# man imapd.conf [root@eg-mailfrontend elsnccpa]# strace -p 7395 Process 7395 attached - interrupt to quit futex(0x7f8af2249e84, FUTEX_WAIT_PRIVATE, 1, NULL^C The frontends running the same cyrus version under Solaris work perfect. Any clue? --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
cyr_expire: IOERROR: reading index header
Hello, running cyrus-2.3.16 under Solaris 10 the nightly event delprune cmd="/opt/cyrus/bin/cyr_expire -E 3" at=0401 emits messages of this kind: Mar 2 04:01:27 cyrus-backend/cyr_expire[5172]: [ID 793632 mail.error] IOERROR: reading index header for user.***: got 0 cache_offset on message 184/21; trying recovery The affected mailboxes have "cyrus.*.NEW" besides the "cyrus.*" files with timestamps of the cyr_expire run. Users report these mailboxes as "demaged" but running reconstruct solves the problem for them. We ran reconstruct for all mailboxes to no effect. Next night the same effect, and every night different mailboxes show this error. The second backend (same config) doesn't show this error. What's the reason? Any pointer towards a solution welcome. --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Error numbers vs. human-readable text
Hello *, what might be the reason for the following problem: Two systems with self-compiled cyrus-2.3.16 (both with Solaris 10) give different responses to IMAP commands. The example is for the "logout" command but all other IMAP commands are also affected. [root@mailbackend-a /]# telnet localhost 143 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. * OK [CAPABILITY IMAP4 IMAP4rev1 LITERAL+ ID MUPDATE=mupdate://mupdate-febe.intern.tu-berlin.de/ AUTH=DIGEST-MD5 SASL-IR] mailcluster2.intern.tu-berlin.de Cyrus IMAP4 (Murder) v2.3.11 server ready . logout * BYE LOGOUT received . OK Completed <--- This looks OK Connection to localhost closed by foreign host. [root@mailbackend-a /]# [elsnccpa@mailbackend-2 ~]$ telnet localhost 143 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. * OK [CAPABILITY IMAP4 IMAP4rev1 LITERAL+ ID MUPDATE=mupdate://mupdate-febe.intern.tu-berlin.de/ AUTH=DIGEST-MD5 SASL-IR COMPRESS=DEFLATE] mailbackend-2-febe.intern.tu-berlin.de Cyrus IMAP Murder v2.3.16 server ready . logout * BYE Unknown code imap 54 . OK Unknown code imap 55 <--- This is ugly Connection to localhost closed by foreign host. [elsnccpa@mailbackend-2 ~]$ What might be the error in compilation or install that prevents correct mapping to the human-readable text? Thanks in advance for any pointer towards a solution, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Re: in mailboxes.db
On Tue, 2 Aug 2011 09:56:19 -0500 Dan White wrote: [ ... ] > For documentation on the mailbox format, see: > > http://www.cyrusimap.org/docs/cyrus-imapd/2.4.10/internal/database-formats.php And exactly this document does not explain the :-( > and > > /imap/mboxlist.h in the source: > > /* flags for types of mailboxes */ > #define MBTYPE_REMOTE (1<<0) /* Not on this server (part is remote host) */ > #define MBTYPE_RESERVE (1<<1) /* Reserved [mupdate/imapd] / > Rename Target [imapd] (part is normal, but > you are not allowed to create this mailbox, > even though it doesn't actually exist */ > #define MBTYPE_NETNEWS (1<<2) /* Netnews Mailbox - NO LONGER USED */ > #define MBTYPE_MOVING (1<<3) /* Mailbox in mid-transfer (part is > remotehost!localpart) */ > #define MBTYPE_DELETED (1<<4) /* Mailbox has been deleted, but not yet > cleaned up */ Thanks. > It appears that mailcluster1 believes a mailbox transfer was initiated, but > has not yet completed (or failed). This happened. > You can use cyr_dbtool to manually edit the entry in your mailboxes.db, or > remove it. See: > > http://asg.andrew.cmu.edu/archive/message.php?mailbox=archive.info-cyrus&msg=52899 One more question: If I use cyr_dbtool for deletion does this interact with the mupdate server which has the correct data? --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
in mailboxes.db
Hallo, which values are possible and what is the corresponding meaning for the in mailboxes.db? After failed XFERs we see - on the murder user.richard^hildebrand 1 mailcluster1 - on mailcluster1 user.richard^hildebrand 8 mailbackend-1-febe!default We would like to know what this means. Is it in error? And if it is, how to solve the problem. --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Re: allowallsubscribe in murder environment
On Mon, 1 Aug 2011 12:16:59 +0200 Frank Elsner wrote: > On Mon, 01 Aug 2011 11:50:52 +0200 Bron Gondwana wrote: > > On Mon, 01 Aug 2011 10:34:25 +0200, Frank Elsner > > wrote: > > > > > > > > Hi folks, > > > > > > in our murder environment we want to allow users to subscribe to > > > mailboxes > > > which are not residing on the server where the users´s mailbox resides. > > [ ... ] > > > But 't' should work fine. Or 1. Or y/yes. > > > > > > >> From the imap.conf example "allowallsubscribe: 0" I took it as numeric > > > but "allowallsubscribe: 1" doesn't work as expected. > > > > > > What have I missed? Any pointer welcome. > > > > Have you set it on both frontends and backends, and restarted everything? > > I did it on the backends only, I'll do it in the frontends also and report. With the same change on the frontends all works as exspected. Thanks for Your help. --Frank Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Re: allowallsubscribe in murder environment
On Mon, 01 Aug 2011 11:50:52 +0200 Bron Gondwana wrote: > On Mon, 01 Aug 2011 10:34:25 +0200, Frank Elsner > wrote: > > > > > Hi folks, > > > > in our murder environment we want to allow users to subscribe to > > mailboxes > > which are not residing on the server where the users´s mailbox resides. [ ... ] > But 't' should work fine. Or 1. Or y/yes. > > > >> From the imap.conf example "allowallsubscribe: 0" I took it as numeric > > but "allowallsubscribe: 1" doesn't work as expected. > > > > What have I missed? Any pointer welcome. > > Have you set it on both frontends and backends, and restarted everything? I did it on the backends only, I'll do it in the frontends also and report. > Did you spell it right? ;) Yes, "allowallsubscribe: 1" --Frank Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
allowallsubscribe in murder environment
Hi folks, in our murder environment we want to allow users to subscribe to mailboxes which are not residing on the server where the users´s mailbox resides. http://www.cyrusimap.org/docs/cyrus-imapd/2.3.16/install-murder.php says | Subscriptions - If users want subscribe to a mailbox other than on their | backend home server, they won't be able to, unless you set | "allowallsubscribe: t" in the backend imapd.confs. This essentially lets any | string be subscribed to successfully. Is it really "allowallsubscribe: t"? ^ >From the imap.conf example "allowallsubscribe: 0" I took it as numeric but "allowallsubscribe: 1" doesn't work as expected. What have I missed? Any pointer welcome. Kind regards, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Re: Upgrading from 2.3 to til 2.4
On Tue, 10 May 2011 12:19:42 -0400 Wesley Craig wrote: > It's usually possible in a murder configuration to do a zero downtime upgrade > by xfer-ing users to a backend running the new version. The admin is then in > control of how much index upgrade load to inflict on the new machine. I'm very interested in details of the "xfer-ing". Can you provide them? --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Move of mailboxes
Hallo ALL, we have 2 backends in a murder environment (running cyrus 2.3.11) which works. And now for the problem: We want to replace both backends by new hardware. Therefore the mailboxes from both must be moved to the replacement machines. Is there a recommended procedure to do the move? Any pointers (even to pitfalls) are welcome. Thanks in advance and kind regards, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
sieve notify not working
Hallo all,
sieve notifications (malto) do not work with this setup (cyrus
2.3.11):
Sieve rule:
# Test: NOTIFY
if header :comparator "i;ascii-casemap" :contains "Subject" "TEST" {
notify :method "mailto" :options "fr...@m28.homelinux.net" :message "You hav
e received a new message
From: $from$
Subject: $subject$
Rule: Test: NOTIFY";
}
"notifyd" daemon ist running and /etc/imapd.conf contains the lines
sievenotifier: mailto
sieve_extensions: fileinto reject vacation imapflags notify include envelope
body relational regex subaddress copy
The mail gets delivered but no notification :-(
Any suggestion for investigation or solution welcome.
Kind regards,
Frank Elsner
Cyrus Home Page: http://www.cyrusimap.org/
List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Re: How to re-sync master/slave?
On Tue, 23 Nov 2010 00:04:52 +1100 Bron Gondwana wrote: > On Mon, Nov 22, 2010 at 11:19:26AM +0100, Frank Elsner wrote: > > > > Hello Cyrus Community, > > > > what is the recommended procedure or best practice to re-sync master/slave > > after failure of synchronization? > > run sync_client -u for each user. I guess I need some more help :-( [r...@mailbackend-a utils]# sync_client −v −u svc-exim Fatal error: No replication mode specified [r...@mailbackend-a utils]# sync_client −v −m svc-exim Fatal error: No replication mode specified [r...@mailbackend-a utils]# sync_client −v −m /mail-pool-a/spool/mail/A/user/svc-exim/ Fatal error: No replication mode specified [r...@mailbackend-a utils]# sync_client −v −m user/svc-exim Fatal error: No replication mode specified [r...@mailbackend-a utils]# sync_client −v −m /mail-pool-a/spool/mail/A/user/svc-exim Fatal error: No replication mode specified [r...@mailbackend-a utils]# From http://www.cyrusimap.org/docs/cyrus-imapd/2.3.16/man/sync_client.8.php I see -r (I dont't want this at the moment) -m and -u What replication mode to specify? Kind regards, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
Re: rsync and cyrusimap.2.4.x
On Tue, 23 Nov 2010 17:13:51 +0100 Martin Pracht wrote: > > Hello, > > with 2.3.X we were happy by having > a standby-system for our mail server, > keeping the data up to date by a > periodical rsync from the working system. Just curious: What do you rsync? The mailboxes only? What else? What about mailboxes.db? --Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
How to re-sync master/slave?
Hello Cyrus Community, what is the recommended procedure or best practice to re-sync master/slave after failure of synchronization? We are using cyrus-2.3.1x and the synchronization seems unstable. What can cause this? Any hints to make it stable? Regards, Frank Elsner Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
