Re: Working Vacation?? [solved?]
Hi, Gary Mills says Sendmail can rewrite header addresses for local delivery. This works well. It's only the envelope recipient that has the domain stripped. And it has only taken a week or two to figure out how to prevent that... cyrusv2.mc patch: -cyrusv2.mc- # divert(-1) # # (C) Copyright 2000 by Carnegie Mellon University # # This sample mc file is for a site that uses the Cyrus IMAP server # exclusively for local mail. This requires Sendmail 8.10 or later. # # $Id: cyrusv2.mc,v 1.4 2001/08/23 23:26:56 leg Exp $ divert(0)dnl VERSIONID(`cyrus v2 sample configuartion') OSTYPE(linux) define(`confBIND_OPTS',`-DNSRCH -DEFNAMES') define(`confTO_IDENT',`0') dnl setting cyrus as the trusted user will make it easier to pass dnl Sendmail's safefile checks. however, it means that someone with the dnl cyrus password could easily become root. dnl define(`confTRUSTED_USER', `cyrus') define(`confLOCAL_MAILER', `cyrus') dnl if you aren't using Sendmail 8.12, you might need to remove dnl the following feature. FEATURE(`preserve_local_plus_detail') FEATURE(`nocanonify') FEATURE(`always_add_domain') MAILER(`local') MAILER(`smtp') MAILER_DEFINITIONS Mcyrus, P=[IPC], F=lsDFMnqA@/:|SmXz, E=\r\n, - S=EnvFromL, R=EnvToL/HdrToL, T=DNS/RFC822/X-Unix, + S=EnvFromL, R=HdrToL, T=DNS/RFC822/X-Unix, A=FILE /var/imap/socket/lmtp LOCAL_RULE_0 Rbb + $+ @ $=w . $#cyrus $: + $1 Dump `EnvToL/' from R= and you get working sieve vacations. The fix turned out to be rather trivial. Does anyone spot problems with this? On Thu, 23 Jan 2003 08:25:05 -0600, Gary Mills [EMAIL PROTECTED] wrote... On Thu, Jan 23, 2003 at 08:02:42PM +0900, Mark Keasling wrote: Sendmail is removing the domain part from smtp envelope TO addresses (RCPT TO) when they are sent to the local deliverer in my case LMTP. Apparently, there is no way to avoid this obnoxious behavior. Hacking on sendmail.cf has had no effect. Yes, sendmail assumes that the domain name is not needed for the envelope recipient for local delivery. No vacation reply gets sent because lmtpd substitutes unspecified-domain for the domain which was removed by sendmail. Its to me check fails because user@unspecified-domain is definitely not in the message's headers. Yes, when I was debugging this, I found that CCing the message to user@unspecified-domain would cause vacation to work! The following patches seem to fix the sieve vacation problem. It would be nice if these patches could be incorporated into the main source so that other people who use sendmail can have vacations too. If someone spots a problem with these patches or notices a deficiency PLEASE let me know how they should be fixed as I'll be putting them into production use RSN. I'm using a similar set of patches. I did submit them to the Cyrus maintainers, but they were rejected as unnecessary. There will be some interactions with aliases where [EMAIL PROTECTED] [EMAIL PROTECTED] and [EMAIL PROTECTED] are all equivalent and all are used. Mail sent to one address will get vacation replies while mail sent to the others will not. I know of no simple way to solve this unless sendmail is kind enough to rewrite the addresses in the message header. Sendmail can rewrite header addresses for local delivery. This works well. It's only the envelope recipient that has the domain stripped. -- -Gary Mills--Unix Support--U of M Academic Computing and Networking- Regards, Mark Keasling [EMAIL PROTECTED]
Re: Working Vacation?? [solved?]
On Thu, 23 Jan 2003, John Alton Tamplin wrote: Mikael Brandström wrote: It should be possible to rewrite rulset Parse0 (and maybe Parse1) to make them leave the local adress qualified while delivering locally. There might even be something among the files in cf/cf which kan fix it, I have a vague memory of something such. Just my 2c. I made changes locally that amounted to always setting $@ whenever $#local was selected as a mailer (for mailers that don't use $h, it shouldn't matter at all). If you redefine the local mailer to be cyrusv2, then look for $#cyrusv2. The places that have to be changed are: Ruleset 0, change: R@$#local $: @ special case error msgs to R@$#local $@ $M $: @special case error msgs (if you aren't masquerading use $j instead of $M -- it could be cleaned up so it would work for more general configs but I was in a hurry to get something working for our setup). In Ruleset Parse1, change: R $+ + $* $+ $#local $: $1 + $2 plussed name? to R $+ + $* @ $+$#local $@ $3 $: $1 + $2plussed name? and then the last two rules in Parse1 change from: R$=L$#local $: @ $1 special local names R$+ $#local $: $1regular local names to R$=L$#local $@ $M $: @ $1special local names R$+ $#local $@ $M $: $1 regular local names Remember that tabs separate the columns, not spaces. Then in the mailer definition, you can pass $u@$h rather than just $u. I am currently using my own deliver program which then connects to LMTP -- this should work for the stock deliver program as well, but for direct LMTP delivery from sendmail you will have to get more creative. I haven't tried it, but it should be possible to copy ruleset EnvFromSMTP to something else and modify it to add @$h to the end of the address if $h is defined, and then change the cyrusv2 mailer to use that for the envelope sender rewrite ruleset. This is what I use (no sendmail m4 hacking neccessary). FEATURE(`mailertable', ... define(`_VIRTUSER_STOP_ONE_LEVEL_RECURSION_', `1')dnl FEATURE(`virtusertable', ... Mcyrus, P=[IPC], F=1lsDFMnqAmXzS, S=EnvFromSMTP/HdrFromSMTP, R=EnvToSMTP, E=\r\n, T=DNS/RFC822/SMTP, A=FILE $h virtusertable: [EMAIL PROTECTED] [EMAIL PROTECTED] mailertable: domain.com lmtp:/var/imap/socket/lmtp Hope this helps. If anyone wants patches to the stock sendmail 8.12.7 proto.m4 file, I will be happy to send them with the usual disclaimers (I have not made any effort to make the changes more general, just working for our particular configuration -- YMMV). I also have changes (including 3 lines of source changes) to allow the w flag to be used with a local mailer yet pass through any users that aren't listed in getpwent (ie, I wanted to have user .forward files yet still allow delivery to shared mailboxes). This hack is even more likely to be a problem in the future, as they are likely to eventually use the flag letter 'y' for something else, but if you want them just ask. One day when I get time I will clean these up and try and submit them to see if they will get included. -- Igor
Re: Working Vacation?? [solved?]
On Fri, 24 Jan 2003, John Alton Tamplin wrote: Igor Brezac wrote: virtusertable: [EMAIL PROTECTED] [EMAIL PROTECTED] So you have to have an entry in virtusertable for every user on the domain? Yes. What is the _mbox%3 part for? userid_mbox is really meant to be a name of the mailbox. %3 handles plussed email addresses, [EMAIL PROTECTED] -- Igor
Re: Working Vacation?? [solved?]
Hi,
Sieve is not sending vacation replies. Why? Here's my stab...
in the dark...
Sieve vacation wants a domain part for the rcpt to address
for its Was the message send directly to me as opposed to
being sent to an alias or got to me some other way check.
Being sent directly to me means my address will appear verbatim
in either the TO, CC or BCC headers. It is comparing the
stmp envelope's TO to the addresses in the message header.
Sendmail is removing the domain part from smtp envelope TO
addresses (RCPT TO) when they are sent to the local deliverer
in my case LMTP. Apparently, there is no way to avoid this
obnoxious behavior. Hacking on sendmail.cf has had no effect.
No vacation reply gets sent because lmtpd substitutes unspecified-domain
for the domain which was removed by sendmail. Its to me check
fails because user@unspecified-domain is definitely not in the
message's headers.
What I did was to:
add parseaddr_getdomain() to lib/parseaddr.c which retrieves
the domain name. It returns unspecified-domain if no domain
name was set.
add parseaddr_setdomain(domain) to lib/parseaddr.c which sets
the domain name.
call parseaddr_setdomain(config_servername) in imap/config.c
after config_servername has been set.
call parseaddr_getdomain() in lib/parseaddr.c and sieve/message.c
before setting unspecified-domain.
The following patches seem to fix the sieve vacation problem.
It would be nice if these patches could be incorporated into
the main source so that other people who use sendmail can have
vacations too. If someone spots a problem with these patches
or notices a deficiency PLEASE let me know how they should be
fixed as I'll be putting them into production use RSN.
There will be some interactions with aliases where [EMAIL PROTECTED]
[EMAIL PROTECTED] and [EMAIL PROTECTED]
are all equivalent and all are used. Mail sent to one address will
get vacation replies while mail sent to the others will not. I
know of no simple way to solve this unless sendmail is kind enough
to rewrite the addresses in the message header.
lib/parseaddr.h
*** parseaddr.h.old Thu Jan 23 18:14:19 2003
--- parseaddr.h Thu Jan 23 17:53:59 2003
***
*** 64,69
--- 64,71
extern void parseaddr_list P((const char *s, struct address **addrp));
extern void parseaddr_free P((struct address *addr));
+ extern void parseaddr_setdomain P((const char *domain));
+ extern char * parseaddr_getdomain ();
#endif /* INCLUDED_PARSEADDR_H */
lib/parseaddr.c
*** parseaddr.c.old Thu Jan 23 18:15:38 2003
--- parseaddr.c Thu Jan 23 17:56:11 2003
***
*** 189,194
--- 189,197
newaddr-mailbox = mailbox;
+ if ( domain !*domain) {
+ domain = parseaddr_getdomain();
+ }
if (domain !*domain) {
domain = parseaddr_unspecified_domain;
}
***
*** 364,367
--- 367,389
return c;
}
}
}
+
+ /*
+ * manage the default domain
+ */
+ static char *parseaddr_domain_name = (char *)0;
+
+ void parseaddr_setdomain (domain)
+ const char * domain ;
+ {
+ if (parseaddr_domain_name) free(parseaddr_domain_name);
+ if (domain *domain) parseaddr_domain_name = xstrdup(domain) ;
+ else parseaddr_domain_name = (char*)0;
+ }
+
+ char * parseaddr_getdomain ()
+ {
+ if (parseaddr_domain_name) return parseaddr_domain_name;
+ return unspecified-domain;
+ }
sieve/message.c
*** message.c.old Thu Jan 23 19:15:02 2003
--- message.c Thu Jan 23 17:16:15 2003
***
*** 460,466
#define U_USER unknown-user
if (a-mailbox || a-domain) {
char *m = a-mailbox ? a-mailbox : U_USER;
! char *d = a-domain ? a-domain : U_DOMAIN;
am-freeme = (char *) xmalloc(strlen(m) + strlen(d) + 2);
--- 460,468
#define U_USER unknown-user
if (a-mailbox || a-domain) {
char *m = a-mailbox ? a-mailbox : U_USER;
! /* char *d = a-domain ? a-domain : U_DOMAIN; */
+ char *d = a-domain ? a-domain : parseaddr_getdomain();
+ if (!d || !*d) d = U_DOMAIN;
am-freeme = (char *) xmalloc(strlen(m) + strlen(d) + 2);
Many thanks to all the busy people, in particular Ken Murchison,
who have provided assistance without recompense. Your accomulated
wisdom pointed me in the right direction. You have my humble thanks
such as they may be.
Regards,
Mark Keasling [EMAIL PROTECTED]
Re: Working Vacation?? [solved?]
On Thu, Jan 23, 2003 at 08:02:42PM +0900, Mark Keasling wrote: Sendmail is removing the domain part from smtp envelope TO addresses (RCPT TO) when they are sent to the local deliverer in my case LMTP. Apparently, there is no way to avoid this obnoxious behavior. Hacking on sendmail.cf has had no effect. Yes, sendmail assumes that the domain name is not needed for the envelope recipient for local delivery. No vacation reply gets sent because lmtpd substitutes unspecified-domain for the domain which was removed by sendmail. Its to me check fails because user@unspecified-domain is definitely not in the message's headers. Yes, when I was debugging this, I found that CCing the message to user@unspecified-domain would cause vacation to work! The following patches seem to fix the sieve vacation problem. It would be nice if these patches could be incorporated into the main source so that other people who use sendmail can have vacations too. If someone spots a problem with these patches or notices a deficiency PLEASE let me know how they should be fixed as I'll be putting them into production use RSN. I'm using a similar set of patches. I did submit them to the Cyrus maintainers, but they were rejected as unnecessary. There will be some interactions with aliases where [EMAIL PROTECTED] [EMAIL PROTECTED] and [EMAIL PROTECTED] are all equivalent and all are used. Mail sent to one address will get vacation replies while mail sent to the others will not. I know of no simple way to solve this unless sendmail is kind enough to rewrite the addresses in the message header. Sendmail can rewrite header addresses for local delivery. This works well. It's only the envelope recipient that has the domain stripped. -- -Gary Mills--Unix Support--U of M Academic Computing and Networking-
Re: Working Vacation?? [solved?]
On Thu, 23 Jan 2003 at 08:25, Gary Mills wrote: On Thu, Jan 23, 2003 at 08:02:42PM +0900, Mark Keasling wrote: Sendmail is removing the domain part from smtp envelope TO addresses (RCPT TO) when they are sent to the local deliverer in my case LMTP. Apparently, there is no way to avoid this obnoxious behavior. Hacking on sendmail.cf has had no effect. Yes, sendmail assumes that the domain name is not needed for the envelope recipient for local delivery. It should be possible to rewrite rulset Parse0 (and maybe Parse1) to make them leave the local adress qualified while delivering locally. There might even be something among the files in cf/cf which kan fix it, I have a vague memory of something such. Just my 2c. Mikael
Re: Working Vacation?? [solved?]
Mikael Brandström wrote: It should be possible to rewrite rulset Parse0 (and maybe Parse1) to make them leave the local adress qualified while delivering locally. There might even be something among the files in cf/cf which kan fix it, I have a vague memory of something such. Just my 2c. I made changes locally that amounted to always setting $@ whenever $#local was selected as a mailer (for mailers that don't use $h, it shouldn't matter at all). If you redefine the local mailer to be cyrusv2, then look for $#cyrusv2. The places that have to be changed are: Ruleset 0, change: R@$#local $: @ special case error msgs to R@$#local $@ $M $: @special case error msgs (if you aren't masquerading use $j instead of $M -- it could be cleaned up so it would work for more general configs but I was in a hurry to get something working for our setup). In Ruleset Parse1, change: R $+ + $* $+ $#local $: $1 + $2 plussed name? to R $+ + $* @ $+$#local $@ $3 $: $1 + $2plussed name? and then the last two rules in Parse1 change from: R$=L$#local $: @ $1 special local names R$+ $#local $: $1regular local names to R$=L$#local $@ $M $: @ $1special local names R$+ $#local $@ $M $: $1 regular local names Remember that tabs separate the columns, not spaces. Then in the mailer definition, you can pass $u@$h rather than just $u. I am currently using my own deliver program which then connects to LMTP -- this should work for the stock deliver program as well, but for direct LMTP delivery from sendmail you will have to get more creative. I haven't tried it, but it should be possible to copy ruleset EnvFromSMTP to something else and modify it to add @$h to the end of the address if $h is defined, and then change the cyrusv2 mailer to use that for the envelope sender rewrite ruleset. If anyone wants patches to the stock sendmail 8.12.7 proto.m4 file, I will be happy to send them with the usual disclaimers (I have not made any effort to make the changes more general, just working for our particular configuration -- YMMV). I also have changes (including 3 lines of source changes) to allow the w flag to be used with a local mailer yet pass through any users that aren't listed in getpwent (ie, I wanted to have user .forward files yet still allow delivery to shared mailboxes). This hack is even more likely to be a problem in the future, as they are likely to eventually use the flag letter 'y' for something else, but if you want them just ask. One day when I get time I will clean these up and try and submit them to see if they will get included. -- John A. Tamplin Unix System Administrator Emory University, School of Public Health +1 404/727-9931
