November 2, 2007 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >
November 2, 2007 [00:00:16] *** meandtheshell has quit IRC[00:01:38] *** jsoftw has left #postfix[00:03:17] <hparker> mark-use: postfixadmin is supposed to send a welcome email when you setup a user... I found a fix in it's *gasp* forums a long time ago for it[00:14:23] *** kgoetz has quit IRC[00:14:35] *** kgoetz has joined #postfix[00:16:44] <kgoetz> this is taking to long :( "Diagnostic-Code: X-Postfix; unknown user: "karl.goetz" ". work already!![00:30:23] *** mark-use has quit IRC[00:39:40] *** pirho has quit IRC[00:42:00] *** pirho has joined #postfix[00:42:29] *** pirho has quit IRC[00:43:55] *** pirho has joined #postfix[00:54:16] *** magyar has joined #postfix[00:57:16] *** theblackbox has quit IRC[01:15:51] *** theblackbox has joined #postfix[01:16:19] <wdp> is it possible to have more than one check_sender_access in smtpd_recipient_restrictions ?[01:17:19] <rob0> You can have as many restrictions as you can keep straight in your head. And probably a few more.[01:17:47] <wdp> hm, that wouldn't be much[01:17:49] <wdp> :p[01:17:53] <rob0> kgoetz, that is not a Postfix error message, or if so, I've never seen it.[01:23:03] <wdp> i read on a postfix FAQ the question "Do people generally have success with setting strict_7bit_headers = yes?" as answer they've written it would be better (if you simply want to block messages with lots of garbage chars in the header) a regexp in header_checks. This one: /[^[:print:]]{7}/ REJECT Your mailer is not RFC 2047 compliant[01:23:24] <wdp> What you think, will this result in false-positives?[01:23:52] <wdp> +to use[01:26:57] <rob0> If you mean for spam reduction, I don't think it would be significant. For my own site, I don't so any content-based approach.[01:28:24] <wdp> well. but is it safe to use that regexp?[01:34:51] *** githogori has quit IRC[01:38:40] *** theblackbox has quit IRC[01:38:59] <rob0> probably[01:39:30] <rob0> you can always test with a WARN action before you do anything drastic[01:40:47] <hparker> rob0: Found me a new career option http://fukung.net/v/150/mcdonalds_hiring.jpg[01:41:57] <rob0> :)[01:42:28] *** ming_zym has joined #postfix[01:50:03] *** havvg has quit IRC[01:53:06] *** oates has quit IRC[01:53:13] *** pitakill has quit IRC[01:58:26] *** sn00p- has joined #postfix[01:59:02] <sn00p-> Does anybody have a postfix guide for freebsd that is up to date? for so people can use outlook and etc?[02:02:12] <rob0> !basic[02:02:12] <knoba> rob0: "basic" : http://www.postfix.org/BASIC_CONFIGURATION_README.html : a good starting place for Postfix beginners, many common questions are answered here.[02:02:25] <rob0> !rob0[02:02:26] <knoba> rob0: "rob0" : a bot that reacts to newly joined users with reciting the !basic factoid :)[02:02:51] <rob0> For virtual mailboxes in FBSD:[02:02:56] <rob0> !ekguide[02:02:57] <knoba> rob0: "ekguide" : http://www.purplehat.org/?page_id=4[02:05:16] <sn00p-> all those guides are out of date[02:05:30] <sn00p-> there at least 3 years old[02:08:20] <hparker> the top guide I know is current for the current version of postfix[02:11:44] <wdp> is it possible to disallow postfix 'sending' except for some domains?[02:13:58] <rob0> You think !basic is out of date? Definitely !rob0 is.[02:15:46] <rob0> IIRC !ekguide is less than a year old.[02:17:16] <rob0> Like all Postfix official documentation, !basic is updated with each release when there are changes.[02:26:35] *** jra has quit IRC[02:36:47] *** pirho has quit IRC[02:47:54] *** amrit|wrk is now known as amrit|afk[03:05:10] *** Tachy_ has joined #postfix[03:05:58] <wdp> warning: dict_nis_init: NIS domain name not set - NIS lookups disabled[03:06:04] <wdp> how can i disable NIS?[03:06:17] *** pitakill has joined #postfix[03:17:48] <rob0> Well, that warning means that NIS is disabled. To do away with the warning, check the default value for alias_maps, set it not to include the NIS lookup.[03:18:53] <wdp> ty.[03:19:53] *** Tachy has quit IRC[03:22:45] <wdp> nini[03:23:23] *** wdp has quit IRC[03:28:26] *** x-ip has joined #postfix[03:39:38] <x-ip> all are sleeping ... hmmm ....[03:39:59] <hparker> no, just not yacking[03:45:01] *** schorpp has joined #postfix[03:51:19] *** felix_da_catz has joined #postfix[03:54:36] *** Motoko-chan has joined #postfix[04:01:33] *** chris_ has joined #postfix[04:11:13] <chris_> Hi, I'm trying to setup a mailhub using a this guide: http://workaround.org/articles/ispmail-etch/#step-7-send-a-test-mail-through-telnet. I'm having trouble testing to see if everything works. When I telnet localhost smtp, and do an ehlo the screen just hangs. I'm using postfix 2.4 from the repo and ubuntu 7.10. I've already checked to make sure everything is configured correctly from the guide. The error log reads: war[04:11:13] <chris_> ning: /usr/lib/postfix/smtpd: bad command startup -- throttling. Haven't been able to get this to work.[04:13:08] *** hing has quit IRC[04:13:27] <hparker> check for typos in your conf files[04:13:36] <hparker> That's my normal problem[04:13:38] <hacim> chris_: look earlier in the log for previous errors[04:14:12] <chris_> hparker: Ok, i'm going back through everything I added now[04:14:31] <chris_> hacim: a couple more, actually now that you mention it[04:14:38] <chris_> postfix/master[7735]: warning: process /usr/lib/postfix/cleanup pid 29983 exit status 1[04:14:48] <chris_> postfix/smtpd[29954]: fatal: open database /etc/postfix/virtual.db: No such file or directory[04:15:16] <hparker> or something like that[04:15:54] <Motoko-chan> There you go.[04:16:08] <Motoko-chan> You need to make virtual.db.[04:16:20] <Motoko-chan> postmap will do that.[04:16:46] <rob0> It involves using a live goat and stump water, and a ceremony at midnight under a full moon.[04:16:57] <rob0> or yeah, maybe postmap[04:17:07] <chris_> I used a hodgepodge of guides to learn postfix. Is virtual.db included in install, i'm pretty sure if i made it, it's no longer necessary[04:17:38] <chris_> I'll check the conf to make sure it's no longer referenced[04:17:45] <Motoko-chan> virtual is included, but you need to use postmap to build it.[04:17:51] <rob0> chris_ set something (virtual_something_maps I bet) to refer to hash:/etc/postfix/virtual[04:18:00] <Motoko-chan> BTW, what is meant by "mailhob"[04:18:03] <Motoko-chan> hub[04:18:08] <Motoko-chan> Like a gateway for scanning?[04:18:51] <chris_> motoko: my terminology could be wrong on this, but i'm trying to setup something w/o any mailboxes just scans/redirects[04:19:08] <chris_> from my understanding that is a mailhub?[04:19:38] <Motoko-chan> I guess that counts.[04:19:45] <Motoko-chan> I'm not up on the terms sometimes[04:20:03] <chris_> sounds like it could make sense :)[04:20:07] <Motoko-chan> http://flakshack.com/anti-spam/[04:20:17] <Motoko-chan> I usually use that as a guide for a scanning gateway.[04:20:31] <Motoko-chan> You can make it deliver to different destinations with a few tweaks.[04:20:48] <chris_> okay thanks, i will check that out next[04:21:22] <Motoko-chan> I don't really follow it as close as I used to. Eventually, one gets their own style.[04:24:33] *** adam__ has joined #postfix[04:27:17] *** cilly has quit IRC[04:27:48] *** cilly has joined #postfix[04:30:43] *** githogori has joined #postfix[04:31:24] *** internat2 has quit IRC[04:31:28] *** internat85 has joined #postfix[04:35:37] *** L|NUX has joined #postfix[04:35:57] *** majikman has quit IRC[04:41:51] <chris_> Okay telnet localhost smtp works now, thanks everyone. I had a couple virtual_* references in /etc/postfix/main.cf that no longer existed, removing those allows me to connect. The test email is not working, but I'll ask if I can't figure it out. Thanks again.[04:53:22] *** x-ip has quit IRC[04:59:55] *** amrit|afk is now known as amrit[05:13:59] *** magyar has quit IRC[05:29:23] *** growltiger has quit IRC[05:33:04] *** pitakill has quit IRC[05:47:29] *** higuita has joined #postfix[06:09:12] *** cilly has quit IRC[06:09:42] *** cilly has joined #postfix[06:14:33] *** kk_CHN has joined #postfix[06:18:01] <kk_CHN> warning: do not list domain mytestdomain.net in BOTH mydestination and virtual_mailbox_domains : why this warning , will it block postfix from its functionality?[06:19:11] <hparker> Pretty obvious, it can only be one type... local or virtual[06:21:07] * Motoko-chan guesses that Postfix won't like it[06:21:23] <kk_CHN> hparker, I folloed a Doc on net , its describing like those steps I followed it . Now the problem is through shell I can send and receive mails for local users .[06:21:43] <hparker> !basic[06:21:44] <knoba> hparker: "basic" : http://www.postfix.org/BASIC_CONFIGURATION_README.html : a good starting place for Postfix beginners, many common questions are answered here.[06:21:53] <hparker> kk_CHN: You might look at ^^^^[06:22:49] <kk_CHN> okay Im looking on that mean while this is my requirement :by using squirrel mail I have to do the same[06:23:00] <kk_CHN> for the local users[06:23:59] <hparker> Anywho, out for the night[06:24:03] *** hparker has quit IRC[06:24:58] *** growltiger has joined #postfix[06:30:51] *** ming_zym has quit IRC[06:33:41] *** ming_zym has joined #postfix[06:41:54] *** amrit is now known as amrit|zzz[06:52:14] *** Edward123 has quit IRC[06:54:43] *** solar_ant has joined #postfix[07:19:34] *** Xen^ has joined #postfix[07:36:59] *** elvedin has quit IRC[07:37:22] <chris_> Still working on setting up a postfix server. I am using this guide: : http://workaround.org/articles/ispmail-etch/#step-7-send-a-test-mail-through-telnet. I'm having trouble setting up an evolution to log into webmaster at example dot com. I have no problem using the command line email browser "mutt" to login into webmaster at example dot com@localhost. I believe the problem has to do with how the webmaster mailbox has been setup. At[07:37:22] <chris_> this point I think the solution lies with DNS/MX settings, but have trouble understanding what the next step is.[07:38:47] *** L|NUX has quit IRC[07:42:04] *** schorpp has quit IRC[07:44:23] *** diveli has quit IRC[07:45:47] *** Motoko-chan has quit IRC[07:58:46] *** L|NUX has joined #postfix[07:59:09] *** ming_zym has left #postfix[08:04:08] *** Zeit|awy has joined #postfix[08:05:00] *** ming_zym has joined #postfix[08:11:24] *** Zeit|idle has quit IRC[08:19:56] *** Xen^ has quit IRC[08:21:24] *** prebur has quit IRC[08:27:15] *** rmayorga has quit IRC[08:29:30] *** Joelwork has joined #postfix[08:29:35] *** veepster has joined #postfix[08:29:40] <veepster> anyone see this?[08:29:41] <veepster> status=deferred (delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]: read timeout)[08:30:18] <Joelwork> hi, what is the difference between the two delays in the logs for postfix... delay=148, delays=0.04/17/0/131[08:32:09] <Joelwork> also say I have around 300M deffered, can I postqueue just 20M?[08:54:16] *** Lap_64 has joined #postfix[08:57:00] *** Tjikkun_work has joined #postfix[09:00:16] *** prebur has joined #postfix[09:01:03] *** war has joined #postfix[09:04:42] *** swarog has left #postfix[09:11:02] *** duki has joined #postfix[09:11:46] *** duki has quit IRC[09:11:55] *** duki has joined #postfix[09:12:46] *** duki has left #postfix[09:13:02] *** duki has joined #postfix[09:23:51] *** rmayorga has joined #postfix[09:26:39] *** r0sk has joined #postfix[09:26:44] <r0sk> Hi[09:27:19] <r0sk> Anyone knows why my postfix is returning good mails as UBE?[09:27:41] <r0sk> is it cause of postfix, clamav or spam-assassin?[09:29:48] <Signum> r0sk: your mail.log will tell you[09:33:53] <r0sk> Signum: cat maillog|grep UBE has no return[09:34:20] <r0sk> (lot of email accounts to see errors with a tail)[09:35:47] <Signum> r0sk: Why do you grep for "UBE"? Can't you search the logs for the sender email address in question?[09:40:19] <r0sk> hmm, I think it's an amavis problem[09:41:36] *** schorpp has joined #postfix[09:51:50] *** phnord has joined #postfix[09:56:59] *** nfi|ermes has quit IRC[09:57:07] <chris_> if you're just setting up virtual mailboxes in postfix and using a program like evolution to check your mail make sure your username is the entire email[09:57:27] <chris_> i just spent a couple hours trying to fix problems with that the above solution[09:58:00] *** rokra_ has joined #postfix[09:58:12] <chris_> The password prompt might even read: user at host dot com@host.com password? <--this works[09:58:31] <r0sk> chris_: yes, the username is the entire email string[09:58:44] <chris_> gosh that couldve saved me a lot of time...[09:58:49] <chris_> live and learn :P[09:59:09] <r0sk> I dont think UBE err will be from postfix[09:59:20] <r0sk> checking clamav and spam-assassin[10:00:24] *** aozturk has joined #postfix[10:02:25] <chris_> r0sk: im setting up a new site with email, in evolution under server for incoming and outgoing mail server is listed as: xyz.com. Is it necessary to have smtp.xyz.com and imap.xyz.com?[10:04:03] *** Edward123 has joined #postfix[10:04:32] <r0sk> that values might be smtp.* or imap.* or an IP, it's not so important I think if they pointed to the right server[10:07:35] <roxlu> hi[10:07:47] <Edward123> morning[10:07:58] <roxlu> I'm trying to send a mail (using an alias) and the other mail server reponds with:Requested action not taken: message refused (in reply to end of DATA command))[10:08:01] <roxlu> what does that mean?[10:09:59] *** rokra has quit IRC[10:20:01] <r0sk> no one with UBE mail errs?[10:20:45] <Edward123> hrm afraid not[10:21:06] <Edward123> roxlu, what refusual do you get when you try to send to our server - server6.netring.co.uk? what's the whole error/rejection message?[10:21:22] <Edward123> you can send to me, edward at netring dot co.uk[10:24:07] <roxlu> Edward123: I think postfix is taking too long to send the end of data character "."[10:28:44] <chris_> webmaster at xyz dot com@local can send and receive mail to any address; webmaster at xyz dot com using evolution can only receive mail. Connection refused by my DNS. Log files mention Relay access denied; I've not been able to send mail out from my virtual mailbox.[10:28:49] *** sn00p- has quit IRC[10:29:06] *** bronb has joined #postfix[10:37:41] *** duki has quit IRC[10:40:55] *** kernel_ has joined #postfix[10:42:02] <kernel_> hey guys, not sure if this is the best place to ask this one. Does AOL added x-spam-flag: yes to some of its email, if so why? In our header_checks we check for this and have had a couple of customers complain just lately they we have rejected them[10:42:29] <kernel_> s/they/that[10:43:24] *** war has quit IRC[10:43:42] *** war has joined #postfix[10:47:51] *** wsxws has joined #postfix[10:47:56] <wsxws> g'day[10:48:13] <wsxws> anybody here knwoing amavis/clamav/spamadssassin ?[10:49:27] <Edward123> i'm not sure about the amavis but i know a fair bit about clamav/sa[10:52:00] <wsxws> i installed those 3 on my system (using amavis as i/o to popstfix)[10:52:07] <wsxws> when i know look at the headers of my emails i can c that those mails have become checked about es but there is no sign of being chackt about spam[10:57:41] *** ming_zym has quit IRC[11:06:27] *** Roobarb-Work has quit IRC[11:13:45] *** L|NUX has quit IRC[11:15:27] *** brancaleone has joined #postfix[11:20:02] *** Twinkletoes has joined #postfix[11:20:09] *** cpm has joined #postfix[11:23:54] *** Zeit|awy is now known as Zeit|idle[11:27:22] *** brancaleone has quit IRC[11:30:19] *** CrummyGummy has joined #postfix[11:32:29] *** brancaleone has joined #postfix[11:33:09] *** lafuma has joined #postfix[11:39:56] *** GutterPunk has joined #postfix[11:41:52] <GutterPunk> Hi. If a multi-homed host resolves it's PTR record to the same hostname for every ip address, is this a faulty configuration? For example, both .30 and .31 point so server.example.net, but server.example.net just points to .30.[11:42:39] <GutterPunk> The RFC1912-2.1 states "For every IP address, there should be a matching PTR record in the in-addr.arpa domain. If a host is multi-homed, (more than one IP address) make sure that all IP addresses have a corresponding PTR record (not just the first one)."[11:43:01] <Trengo> why would you have two ptrs resolving to the same name?[11:45:45] <lafuma> good morning[11:46:04] <lafuma> could anyone help me out with setting up a mail server ?[11:46:05] <cpm> Trengo, that question is giving me a headache[11:46:18] <GutterPunk> Because I didn't think it would matter. Is it better policy to just use the ip1., ip2.host.example.net then?[11:46:26] <cpm> lafuma, read the topic please, and pay careful attention to the getting help link[11:46:28] <lafuma> i've set up postfix and dovecot, but i cant get the damn thing working properly[11:46:53] <lafuma> sec[11:47:14] <lafuma> ahha, i see[11:47:15] <Trengo> cpm im sorry, do you want a paracetamol?[11:47:30] <cpm> Trengo, umm, maybe[11:47:53] * Trengo dccs 2 500mg paracetamol pills to cpm[11:47:56] <lafuma> i get this error when trying to send email to my postfix server:[11:48:06] <lafuma> Nov 2 12:28:30 void postfix/smtpd[17973]: NOQUEUE: reject: RCPT from smtp-out.neti.ee[194.126.126.39]: 554 5.7.1 <marek at webchat dot ee>: Relay access denied; from=<marek at trekker dot ee> to=<marek at webchat dot ee> proto=ESMTP helo=<smtp-out.neti.ee>[11:48:16] <GutterPunk> We've only experienced this issue sending mail from a server to one domain, which apparently checks the ptr first, and which afterwards checks if the ip of that host is the same as the one the connection initially came from. But I don't think the RFC's require that[11:48:17] <cpm> Anyway, if the PTR records point at different subnets, then it makes sense of course. In a split horizon environment. Where mail.domain.tld outside is on IP, and mail.domain.tld inside is another.[11:48:30] <lafuma> webchat.ee sits in my server then[11:48:41] <cpm> However, it makes no sense to have multiple IPs on the same subnet pointing to the same A record. That's wasteful[11:49:02] <Trengo> btw whats cpm? Critical Path Method?[11:49:47] <cpm> umm, initials[11:49:51] *** brancaleone has quit IRC[11:50:15] <Trengo> oh right[11:50:38] <Trengo> dang you're lucky, your initials make up a nice nick[11:50:48] <lafuma> any ideas on my problem ?[11:51:19] <GutterPunk> cpm: I thought it would be wasteful to go and specify an A record for each ip of a multi-homed host[11:52:48] <GutterPunk> So what would you say is the best practise in this case? And is my configuration actually faulty? Should a mail server be blocking mails from my host?[11:52:49] <cpm> :)[11:52:49] <cpm> lafuma[11:52:50] <cpm> !basic[11:52:51] <knoba> cpm: "basic" : http://www.postfix.org/BASIC_CONFIGURATION_README.html : a good starting place for Postfix beginners, many common questions are answered here.[11:53:28] <cpm> GutterPunk, are the IPs on the same or different subnets?[11:53:42] <GutterPunk> cpm, same subnet[11:53:47] <cpm> then why have them?[11:53:57] <cpm> same subnet, one IP is sufficient[11:54:46] <GutterPunk> cpm: well, it's multi-homed because it's a webserver which hosts pages for some domains using https.[11:55:28] <wsxws> hm[11:55:30] <cpm> Okay, that makes sense. So bind your mail server to 1 IP address (the IP address of the physical network device would be a good idea)[11:55:42] <wsxws> when i miss an x in the permissions of a path, how to ad that ?[11:55:44] *** ek has quit IRC[11:55:46] <cpm> and you're golden[11:56:16] *** brancaleone has joined #postfix[11:56:18] <cpm> btw, putting all your eggs in one basket like that, is less than optimal[11:57:41] <GutterPunk> cpm: well, the primary address is .30. The smtp service is running using the .31 address. But when freenet's mail server gets mail from this host, it checks the PTR to see to which one it points, finds host.example.net. But host.example.net resolves to .30[11:58:20] <GutterPunk> And since the connection came from .31, it apparently rejects it and happily points me to RFC 1912, 2.1. :)[11:58:25] *** cilly has quit IRC[11:59:29] <GutterPunk> Well I will make it run using the primary IP now. It should work fine then.[12:00:41] <GutterPunk> But still, I don't think this is required by the RFC, and was just wondering what you guys thought. But it is ok right, to have many ip's point back to the same PTR record? Otherwise, I've got quite some changes to make.[12:00:57] *** wsxws has quit IRC[12:01:51] <cpm> No, it is NOT OKAY to point multiple PTRs at the same IP.[12:02:10] <f3ew> one PTR record, one IP[12:02:17] <f3ew> one IP, one PTR record[12:02:21] <cpm> it won't work either.[12:02:54] *** havvg has joined #postfix[12:03:43] <Trengo> ips pointing at the same name?[12:03:45] <Trengo> why?[12:04:08] <cpm> GutterPunk, check option smtp_bind_address[12:04:47] <cpm> Trengo, it is a violation of the RFC for one thing, for another thing, that's not how dns[12:05:20] <cpm> GutterPunk, smtp_bind_address = ipaddressofyourmailhost[12:05:38] <Trengo> cpm i agree entirely, and i cant see any use for it[12:06:02] <cpm> There is no use for it. And the just hearing the question scares me.[12:06:03] <cpm> :)[12:06:15] <Trengo> having a name resolve to several IPs isnt good, can be used for round robin DNS load balancing, but sucks[12:06:51] <Trengo> but an IP resolving to several names sounds just nasty[12:06:59] <cpm> it is. That will break things[12:07:02] <GutterPunk> the name only resolves to 1 ip, BUT many PTR's point back to the same A record[12:07:13] <GutterPunk> so it's the other way around :)[12:07:32] <cpm> GutterPunk, no.[12:07:40] <cpm> 1 IP, 1 PTR[12:07:57] *** cilly has joined #postfix[12:08:06] <GutterPunk> alright, every ip has a ptr as well[12:08:11] <GutterPunk> and only one[12:08:13] <cpm> you can have multiple A records to one IP[12:08:19] <GutterPunk> ah[12:08:40] <Bejgli> how can i make a mysql virtual catch-all address for postmaster@* ? i tried adding postmaster@ into alias table with goto value: postmaster@somedomain but postfix gives me user unknown[12:11:21] *** brancaleone has quit IRC[12:12:41] *** brancaleone has joined #postfix[12:13:22] *** schorpp has quit IRC[12:24:04] <internat85> just type the username postmaster i think[12:24:07] <internat85> dont hold me to that though[12:24:47] <internat85> so the source would be "postmaster", and the destination would be "email at enddomain dot com"[12:24:50] *** mbainter has left #postfix[12:25:27] <cpm> Bejgli, if you want to catch all the mail for postmaster, that is NOT a catchall. do you want to do a catch all and direct it to postmaster? or do you just want postmaster@ all your domains to go to one place?[12:25:34] <cpm> In either case,[12:25:37] <cpm> read the docs plesae.[12:25:41] <cpm> please.[12:25:48] <cpm> http://www.postfix.org/VIRTUAL_README.html[12:25:56] <Bejgli> cpm: just want postmaster@ all your domains to go to one place?[12:25:57] <cpm> !virtual[12:25:57] <knoba> cpm: "virtual" : a way to configure additional domains and user accounts (that do not need to exist in your /etc/passwd). See: http://www.postfix.org/VIRTUAL_README.html[12:26:10] <Bejgli> but in mysql virtual, i couldnt find this in virtual_readme[12:26:29] <Bejgli> because even though i add postmaster@ goto: postmaster@domain it doesnt work[12:26:50] <cpm> that is the wrong syntax. It's just an alias, as shown on that page.[12:27:11] <cpm> postmaster at domain1 dot tld postmaster[12:27:14] <cpm> postmaster at domain2 dot tld postmaster[12:27:20] <cpm> and so on, in a hash table[12:27:44] <cpm> and point virtual_alias_maps at that hash table[12:27:53] <cpm> so if you have a mysql lookup,[12:27:56] *** tmjb has joined #postfix[12:28:07] <Bejgli> but isnt it possible to have only 1 line for postmaster@* and i wouldn't need to add for each domain ?[12:28:13] <cpm> you would do something like virtual_alias_maps = hash:/path/to/virtual mysql:/path/to/query.cf[12:28:48] <tmjb> how to delay or stop some large mail from mailq i try postsuper -h queid but noluck ?[12:29:03] <cpm> Bejgli, that is not what the docs say, is it?[12:29:32] <cpm> Bejgli, you tried it and it didn't work. Did it?[12:29:37] <Bejgli> docs dont say about this particular case[12:30:09] <cpm> How so?[12:30:12] *** Roobarb-Work has joined #postfix[12:30:25] <Bejgli> postfix only looks up postmaster@domain and then @domain , it doesnt care if i have postmaster@ in my alias table[12:30:48] <cpm> As well it shouldn't. since you have a default postmaster alias, it would loop[12:30:50] <Trengo> is that a mysql query?[12:31:00] <Bejgli> Trengo: yes[12:31:01] <cpm> Bejgli, why don't you just do it like the docs say?[12:31:07] <Trengo> Bejgli play with it[12:31:18] <cpm> Do you have a default postmaster alias?[12:31:35] <cpm> for the parent?[12:31:43] <Bejgli> no[12:31:46] <cpm> k[12:31:58] <Bejgli> how can i set it?[12:32:36] <cpm> might work. But since you are going outside the docs, you are on your own. Like Trengo said, play with it. I don't like the syntax you're using, and I think it's going to cause problems, but who cares what I think[12:32:37] <cpm> :)[12:32:56] <Bejgli> but what do u see in docs, please[12:33:07] <Bejgli> i cant find that particular section[12:33:21] <Bejgli> where it says about aliasing for user@* for every domain[12:33:55] * f3ew hands Bejgli a regular expression[12:34:13] <f3ew> user@ Matches all mail addresses with the specified user part.[12:34:18] <f3ew> From access(5)[12:34:27] <Bejgli> but i cant use it in mysql[12:34:34] <f3ew> Uh?[12:34:47] <Bejgli> query = SELECT goto FROM alias WHERE address='%s' AND active = 1[12:34:55] <Trengo> liek[12:35:02] <Trengo> like[12:35:03] <Bejgli> alias='postmaster@' , goto='postmaster@domain'[12:35:09] <Trengo> dude[12:35:19] <Bejgli> address like '%s' ?[12:35:22] <f3ew> no[12:35:33] <f3ew> Can you check using postmap -q?[12:35:38] <Bejgli> yes[12:35:41] <f3ew> What version of Postfix are you using?[12:36:05] <Bejgli> 2.3.6[12:36:48] *** lafuma has quit IRC[12:36:57] <f3ew> oh wait, you are looking at alias_maps, not access?[12:37:05] <f3ew> or virtual_alias_maps[12:37:13] <Bejgli> yes, virtual_alias_maps[12:37:20] <f3ew> Try just postmaster[12:37:22] <f3ew> no @[12:37:56] <Bejgli> Recipient address rejected: User unknown[12:38:24] <Bejgli> i can turn on verbose logging for smtpd if needed[12:39:34] *** Lukemob has joined #postfix[12:40:03] <Lukemob> hello[12:41:08] <Lukemob> do I need to have created a system user to have a mail user? ... isn't possible to use postfix with virtual users?[12:41:38] <Bejgli> http://www.postfix.org/VIRTUAL_README.html[12:41:49] <Lukemob> thx[12:42:22] <f3ew> Bejgli just turn on mysql query logging[12:42:27] <f3ew> and see what queries actually run[12:42:51] <f3ew> postmap -q postmaster at example dot com mysql:/... should return the correct result[12:42:58] <f3ew> !virtual[12:42:59] <knoba> f3ew: "virtual" : a way to configure additional domains and user accounts (that do not need to exist in your /etc/passwd). See: http://www.postfix.org/VIRTUAL_README.html[12:43:43] * cpm really thinks that postmaster@ and abuse@ needs to be handled intentionally[12:44:03] <cpm> but I also thinks I could very well be wrong[12:44:07] *** wdp has joined #postfix[12:44:14] <Trengo> no, you're very right[12:44:37] <Trengo> unfortunately, postmaster gets too much junk due to misconfigured servers across the net[12:44:38] <cpm> Trengo, I handle all my postmaster@ and abuse@ with mailing lists.[12:44:50] <cpm> I do this so I can handle hundreds of domains on purpose[12:45:14] <cpm> so, aside from the domains that I personally host, I handle postmaster and abuse for a lot of folks I know.[12:45:27] <cpm> and it's been a long time since I've seen a 'real' postmaster email come through.[12:45:46] <Trengo> whats a real postmaster email?[12:45:47] <cpm> so, maybe a catchall isn't such a bad idea.[12:45:52] <Lukemob> can I use also a subdomain as a mail domain?[12:45:53] <Lukemob> like[12:45:56] <cpm> Trengo, someone actually trying to contact the postmaster[12:45:59] <Lukemob> lukemob at me dot lukemob.com[12:46:02] <Bejgli> why should i make lots of alias making postmaster@domain1 -> postmaster@maindomain , postmaster@domain2 -> postmaster@maindomain, ... when i could have postmaster@* -> postmaster@maindomain ?[12:46:28] <Trengo> cpm oh happens sometimes[12:46:31] <Trengo> even i do it[12:46:33] <cpm> Bejgli, it's a reasonable question[12:46:36] <wdp> cpm, in 10 years, i got only one 'real' mail to a abuse@ address[12:46:44] <wdp> ;-)[12:46:58] <Trengo> i get lots of mail to abuse[12:47:36] <cpm> wdp, I haven't seen any real mail to abuse at in a very long time. But postmaster occasionally gets valid traffic. But not very much.[12:47:46] <cpm> I think that's the problem.[12:48:16] <cpm> I wouldn't be opposed to delisting domains that don't respond to email sent to those addresses[12:48:19] <cpm> :)[12:49:13] <cpm> but, I ain't aol either. I can't imagine the pain[12:49:41] <Trengo> they got the money, they bloody well should have the budget[12:49:46] <cpm> Trengo, agreed[12:50:01] <Trengo> need 50 guys to look at postmaster? hire them![12:50:07] <cpm> but there is an inverse function[12:50:14] <cpm> 500? Hire them.[12:50:23] <Trengo> need better systems? build or buy them[12:50:30] <cpm> as far back as 94, postmaster at ibm dot com became a black hole.[12:51:34] <wdp> someone should code a virus, destroying a spammers box ;)[12:51:43] <Bejgli> f3ew: [postfix/smtpd] maps_find: virtual_alias_maps: postmaster@domain: not found[12:51:56] <Bejgli> f3ew: [postfix/smtpd] maps_find: virtual_alias_maps: @domain: not found[12:51:57] * f3ew sighs[12:52:10] <f3ew> Query logging[12:52:15] <Bejgli> oh, sorry[12:52:24] <Trengo> wdp there's no such thing im afraid[12:53:04] <f3ew> Trengo, postmaster@ needs good people, and there is a known shortage of good people[12:53:39] <Trengo> f3ew agreed, but thats not a valid reason for throwing mail away[12:53:57] <f3ew> ah, agreed[12:53:58] <Trengo> "we cant handle it because we cant hire good enough staff"[12:54:13] <cpm> I had a business idea a few years ago, for outsourcing postmaster and abuse. why I set things up like I did.[12:54:22] <cpm> However, I think I was about 5 years too late.[12:55:31] <cpm> the idea was to have clients point all their postmaster/abuse at my mailing lists, and hire f3ew to read'em.[12:55:32] <cpm> :)[12:55:38] <f3ew> heh[12:55:44] <cpm> I'm not really kidding[12:55:49] <f3ew> aieeeeeee[12:56:00] <cpm> however, I think things are past the point of caring[12:56:08] <f3ew> mostly, yes[12:56:22] <f3ew> OTOH, spammers are now moving to HTTP, so email will become usable again[12:56:32] <Trengo> they are?[12:56:34] <Lukemob> xD[12:56:36] <cpm> when all the fuss started up over spf, I though the concept was resurectable[12:57:20] <cpm> now, however, folks want to get rid of whois records altogether, even further distancing responsible parties from domain records.[12:58:03] <Trengo> he[12:58:10] <Trengo> spammers do, right?[12:58:19] <cpm> yes, spammers do.[12:58:20] <cpm> :)[12:58:32] <Bejgli> f3ew: http://rafb.net/p/875XSU51.html please have a look, its the full querylog of mysql when trying to send mail for postmaster@somedomain and having 'postmaster' in my alias table[12:59:58] <cpm> Trengo, it's my opinion, that as much fuss as folks make over 'their' domains, ultimately, they don't actually care.[13:00:16] <cpm> when role accounts started showing up, , , uhh, in 96?[13:01:14] <cpm> that kinda said it right there. Where a role account is a fair idea for a 'real' company, practice has shown that in reality, it isn't valid at all.[13:02:24] <cpm> This year, I've had about a dozen occasions to pick up the phone and dial up the tech contact for a domain, and every time, I got a real person.[13:02:29] <cpm> that will be going away soon.[13:02:45] <wdp> cpm, so and one dollar per mail?[13:02:52] <Trengo> cpm you're quite right, people complain but dont want to do any actual work[13:03:03] <f3ew> My answer to that is fairly simple[13:03:06] <f3ew> block and forget[13:03:13] <wdp> cpm, if you get 10 000 spam mails daily (like i) to the postmaster address, you would be rich after a week :p[13:03:16] <cpm> f3ew, a good concept.[13:03:31] <cpm> wdp, that's why you hire f3ew,[13:03:39] <f3ew> user address, address, ...[13:03:40] <f3ew> Redirect mail for user@site to address when site is equal to $myorigin, when site is listed in $mydestination, or when[13:03:40] <f3ew> it is listed in $inet_interfaces or $proxy_interfaces.[13:03:42] <f3ew> Hmmmmm[13:04:05] <f3ew> Bejgli why not just use a regexp?[13:04:13] <wdp> <cpm> when all the fuss started up over spf, I though the concept was resurectable[13:04:19] <wdp> i like spf. but there are some problems.[13:04:24] <cpm> I don't like spf.[13:04:39] <cpm> I thought I did when the idea started getting kicked around.[13:04:44] <Bejgli> f3ew: im already using mysql for a lot of domains and it seemed obvious to have that particular postmaster alias in mysql as well[13:05:01] <cpm> but breaking forwarding is too high a price to pay.[13:05:10] <cpm> in my insane world view[13:05:15] <Bejgli> f3ew: it'd be another lookup and thats more overhead[13:05:19] <cpm> I like domain keys.[13:05:28] <f3ew> Not really much[13:05:37] <cpm> however, until concepts like ca-cert become fully viable, I think that's off the table also[13:05:57] <cpm> and, if folks actually trusted them, imagine the hell that would issue forth[13:09:05] <Bejgli> f3ew: so , do u recommend to have both a regexp and a mysql lookup map for aliases and have the postmaster@* in the regexp table?[13:10:18] *** wsxws has joined #postfix[13:10:34] <wsxws> g'day again :)[13:12:27] <wsxws> i run a webserver with debian etch, i also installed postfix/cyrus/pam.d/postgrey/amavis/clamav and spamassassin. everything seems to work fine meanwhile except the spammassassin. when i look at recieved emails there is nothing to c about spammlevels[13:14:23] <f3ew> yes[13:15:05] <wsxws> any suggestions where to look for errors or which settings could be wrong ?[13:16:10] <Bejgli> wsxws: try sending a spam-test email, a mail with this string in its body: XJS*C4JDBQADN1.NSBN3*2IDNEN*GTUBE-STANDARD-ANTI-UBE-TEST-EMAIL*C.34X[13:16:46] <wsxws> a sec ill try[13:18:00] <wsxws> hm, this email does not reach me[13:19:28] <wsxws> that would mean that spamassassin works[13:19:49] <wsxws> then it would be intresting why it does not mark the other mails with spamlevels[13:20:23] <Bejgli> it marks them[13:20:28] <Bejgli> but only above a certain level[13:20:48] <wsxws> hm[13:20:50] <wsxws> look[13:20:57] <Bejgli> wsxws: if u r using amavisd, search for $sa_tag_level_deflt[13:21:41] <wsxws> $sa_tag_level_deflt = 0; # add spam info headers if at, or above that level[13:21:54] *** rokra_ has quit IRC[13:22:01] *** Lukemob has quit IRC[13:22:08] <Bejgli> wsxws: maybe your mails get a spam level under 0, have a look at your amavisd.log or syslog (if amavisd logs there)[13:22:22] <Bejgli> wsxws: and search for Hits: X (where X is the spam level)[13:22:28] <wsxws> okies ill set it to -1000[13:22:40] <Bejgli> wsxws: afaik -999 is the lowest possible[13:23:16] <wsxws> ah[13:24:51] <wsxws> hm, now i set it to -999 and dind a /etc/init.d/amavis restart[13:25:10] <wsxws> shouldnt it be that there is a spamheader in every email now ?[13:25:52] <Bejgli> it should[13:26:04] <Bejgli> if it goes through spam filters[13:27:08] <wsxws> it goes[13:27:23] <wsxws> when i have a look at the syslog there a lots of those hits entries[13:27:48] <wsxws> but why doenst it add the headers to the emails ?[13:28:51] *** Zeit|idle has quit IRC[13:28:58] *** bronb has quit IRC[13:29:26] <wsxws> i also found your testemail: Hits: 1001.636[13:29:27] <wsxws> wow[13:29:34] <wsxws> cool testing body :P[13:29:57] <Bejgli> wsxws: http://gentoo-wiki.com/HOWTO_Email:_A_Complete_Virtual_System_-_Installing_Amavisd_and_SpamAssassin[13:33:02] <wsxws> hm[13:33:07] <wsxws> i dont have any amavisd.conf[13:33:10] <sysmonk> wsxws: what's the test email?:))[13:33:39] <sysmonk> never saw more than ~80 hits[13:33:40] <wsxws> sysmonk just put the following into the body: XJS*C4JDBQADN1.NSBN3*2IDNEN*GTUBE-STANDARD-ANTI-UBE-TEST-EMAIL*C.34X[13:33:49] <sysmonk> ah[13:34:01] *** kk_CHN has quit IRC[13:34:53] <wsxws> i mean my spammassassin hasngot got much to to do anyway, i have 6 RBL's and postgrey before it goes to spamassassin, but i really would like it would mark my headers[13:35:52] <wsxws> http://wsxws.ws/mail.cgi <<<-- but those spammers didnt let me any choice[13:36:32] <sysmonk> and? what's bad here ? :)[13:36:33] <wsxws> but back to my headerproblem, that howto is totaly different to my config-files[13:36:56] <wsxws> i still need to know why my working spamassassin doesnt not mark my headers[13:37:13] <wsxws> could it be a problem of rights to write into the headers ?[13:37:59] <wsxws> spamd is running under user root atm[13:42:35] <wsxws> hm[13:42:38] <wsxws> nobody in here nomore ?[13:45:28] *** x-ip has joined #postfix[13:47:59] <Twinkletoes> I am but I don't knwo much[13:48:47] <wsxws> hehe, thanks anyway Twinkletoes :)[13:49:53] <Twinkletoes> When Postfix and Dovecot are both installed, is it Postfix which receives mail from another SMTP server and also places the mail in the relevant users mail storage, or doe sit hand off to Dovecot for that?[13:51:43] <Trengo> Twinkletoes depends what you tell it to do[13:51:52] <Trengo> default is postfix does the local delivery[13:51:59] * cpm hands off to Trengo[13:52:10] <Twinkletoes> Trengo: Ah, ok - thank you. I@ll leave it that way if I can then[13:52:52] <wsxws> Trengo do u have any clue why my spamassassin might not mark the emailheaders ?[13:52:55] <Trengo> cpm see, i've been reading you :)[13:53:44] *** TestMaster has quit IRC[13:53:56] <f3ew> wsxws, you don't need spamd w/ amavisd[13:54:17] <wsxws> ah, what i might forgot to tell you, amavis marks the headers according to the viruschecks, just the spampart is missing[13:54:52] <Twinkletoes> From what I can tell, Postfix (as default) will deliver to '~<user>/Maildir/' or '~<user>/Mailbox'. I always thought mail would be delivered to /var/spool/mail or equivalent. Was my thought wrong?[13:54:56] <f3ew> wsxws, you have ti tell amavisd the score at which to start marking[13:55:15] <f3ew> Twinkletoes depends on the setting on home_mailbox[13:55:45] <wsxws> whosoever is filtering the spam atm, in syslog i can see that spamfiltering works, i can c all those little hits and the higher one from the testmails. the lower levels are delivered, the higher arent, fine. just the headers arent marked[13:55:50] <Twinkletoes> f3ew: home_mailbox just gives the choice of Mailbox or Maildir/, it doesn't seem to affect which directory to place it in[13:56:22] <f3ew> Twinkletoes comment it out[13:56:41] <wsxws> f3ew $sa_tag_level_deflt = -999; # add spam info headers if at, or above that level that my entry atm (before i had set it to 0 but i did not c any changes at the headers neither)[13:57:25] <f3ew> Hmmmmm[13:57:37] <Twinkletoes> f3ew: it says "The home_mailbox parameter specifies the optional pathname of a mailbox file _relative to a user's home directory_". I assume Postfix will auto-detect the style of mailbox storage. I always used to think mail went to /var/spool/mail, or is that jsut for undelievered mail?[13:58:46] <f3ew> Twinkletoes it goes to /var/spool/mail if home_mailbox is not set[13:59:08] <Twinkletoes> f3ew: Ah, than you - that explaions a lot :)[14:00:30] *** Danielss89 has joined #postfix[14:00:31] <Danielss89> hi[14:00:40] <f3ew> lo[14:04:54] <wsxws> f3ew meanwhile i added the line $sa_spam_report_header = 1; but it still does not mark[14:05:49] <Trengo> wsxws sorry, no idea, im not used to amavis[14:06:41] *** Joelwork has quit IRC[14:06:44] *** Joelwork has joined #postfix[14:06:55] *** apeiron has quit IRC[14:07:12] <wsxws> that really becomes annoying, i mean it filters spam, i can c it in the logs, it cant be a permission thingie, cause amavis is allowed to write marks into the headers like i can c at the clamav entry[14:07:20] <wsxws> so what the hell can that be ?[14:08:50] <f3ew> dunno[14:10:02] <wsxws> i set up the whole system manually, everything is working fine meanwhile, it cant be that i cant fix that last little thingie[14:15:35] *** eikke_ has quit IRC[14:17:43] *** jelly has quit IRC[14:18:06] *** eikke has joined #postfix[14:18:26] *** jelly has joined #postfix[14:19:14] *** jelly has quit IRC[14:19:58] *** jelly has joined #postfix[14:23:37] <Trengo> wsxws did you look at /etc/mail/spamassassin/local.cf?[14:25:00] <wsxws> not yet i was looking for a .conf :P[14:27:13] <wsxws> hm[14:27:25] <wsxws> those settings all cann be fount in the amavis.conf[14:28:44] <tmjb> f3ew,I used amavisd-new but mailscanner is much better (my expirience) what is your opinion ?[14:29:03] <CrummyGummy> Hi all, I've got a wierd postfix (I assume) problem.[14:29:17] *** PierreTramo has quit IRC[14:29:24] <CrummyGummy> My logs say the following:[14:29:32] <CrummyGummy> delivery temporarily suspended: Host or domain name not found. Name service error for name=flashmedia.co.za type=MX: Host not found, try again[14:29:57] <CrummyGummy> But I can ping the MX host for that domain. Any ideas what I'm missing.[14:31:16] *** hark has quit IRC[14:31:21] <wsxws> hm, editing the local.cf does not make any changes[14:31:26] <CrummyGummy> And telnet into it on port 25.[14:31:46] <CrummyGummy> wsxws: Looking...[14:32:22] *** vms has joined #postfix[14:32:25] <CrummyGummy> I have no such file...[14:32:30] *** memetic has quit IRC[14:32:34] <jpalmer> CrummyGummy: dig mx flashmedia.co.za[14:33:01] <CrummyGummy> mx.flashmedia.co.za. 0 IN A 208.69.32.137[14:33:15] *** poisdeux has joined #postfix[14:33:27] <wsxws> looking ... ?[14:33:28] <jpalmer> that's an A record, you're looking for an MX record.[14:34:03] <CrummyGummy> dig flashmedia.co.za mx[14:34:08] <CrummyGummy> flashmedia.co.za. 86188 IN MX 10 mx3.lantic.net.[14:34:38] <jpalmer> CrummyGummy: and you did that from your postfix machine?[14:34:45] <CrummyGummy> wsxws: for local.cf, may have got some wires crossed.[14:34:48] <CrummyGummy> jpalmer: Yup[14:35:13] <jpalmer> CrummyGummy: sounds like a transient DNS issue[14:35:33] <jpalmer> postfix can't find the MX for that domain, but your box DNS libraries can[14:35:51] <wsxws> CrummyGummy using amavis the local.cf files are completely marked out[14:36:28] <jpalmer> wsxws / CrummyGummy you've crossed wires. you two aren't talking about the same things.[14:36:49] <CrummyGummy> hehe, agreed.[14:37:27] <CrummyGummy> jpalmer: Is there a setting where I can fix this? Its been going on since the 31st.[14:37:30] *** jelly has quit IRC[14:37:57] <wsxws> dunno[14:38:06] *** jelly has joined #postfix[14:38:22] <jpalmer> CrummyGummy: that'd be beyond my knowledge level. I can identify the problem, but wouldn't know how to fix it[14:38:23] <wsxws> i start to hate that thingie[14:38:38] <CrummyGummy> jpalmer: Thanks[14:41:04] <cpm> mx3.lantic.net is the MX record for flashmedia.co.za[14:41:07] <cpm> what is the problem?[14:41:52] <CrummyGummy> cpm: The problem is that mail isn't being delivered with this error.[14:41:57] <cpm> k[14:42:00] <cpm> don't paste it.[14:42:06] <cpm> is your postfix running chroot?[14:42:13] <CrummyGummy> nope[14:42:32] <cpm> how do you know? your chroot column in master.cf set to 'n' ?[14:43:55] <CrummyGummy> Its n in qmgr,local,virtual,maildrop,uucp,ifmail,bsmtyp,scalemail and mailman. Its a - in the rest.[14:44:12] *** f3ew has quit IRC[14:45:01] <cpm> change that '-' in smtp to 'n', then STOP postfix, the START postfix (don't do a reload)[14:45:08] <cpm> and see if that makes a difference.[14:45:14] <cpm> If so, then *something* changed.[14:45:29] <cpm> does your machine auto-update by any chance?[14:46:07] <CrummyGummy> I do it manually but I havn't had a manual config file change in ages.[14:46:30] <cpm> does your machine auto-update by any chance?[14:46:41] <cpm> hint (y/n)[14:47:15] <CrummyGummy> n[14:47:33] <cpm> k[14:47:34] <cpm> good[14:47:47] <CrummyGummy> Cool, that worked. Thanks.[14:47:58] <cpm> wierd. So, *something* changed[14:48:02] <wsxws> at least one got helped[14:48:07] <cpm> I'm glad it worked, but you were running chroot,[14:48:19] <cpm> which was just fine before.[14:48:25] <cpm> so, something in your chroot changed.[14:48:39] <cpm> did you upgrade anything on the 31st? named maybe?[14:49:15] <CrummyGummy> No, we changed hosting environments though. I had to change IP addresses and a few other things.[14:49:28] <cpm> ah[14:49:35] <cpm> I'd call that a pretty major change.[14:49:40] *** eikke has quit IRC[14:49:51] <cpm> Well, keep poking at it, at least the mail is going through. But I think you need to revisit and repair the chroot[14:50:44] <CrummyGummy> cpm: Thanks.[14:52:31] <wsxws> my spamassassin is still not marking the headers of my email (im using amavis, marks for x-virus are there, spamassassin is filtering according to syslog, default-level fpr marking is at -999 in amavis conf)[14:59:08] *** aozturk has quit IRC[14:59:19] *** aozturk has joined #postfix[15:02:20] *** GutterPunk has quit IRC[15:03:04] <wdp> rob0 alive?[15:03:45] <wsxws> dunnp, but i could offer u a nice spammassassin/amavis-prob for your waiting time :P[15:04:27] <wdp> if i have a gateway for incoming mail, another mailserver recieves the mail instead of the gateway, can i easily 'relay' such mails from that mailserver to the gateway, and will the gateway then resend it to that mailserver, or will that result in loops back to myself ?[15:04:52] <wdp> wsxws, yeah, i think i can help[15:05:46] <wsxws> my spamassassin is not marking the headers of my email (im using amavis, marks for x-virus are there, spamassassin is filtering according to syslog, default-level fpr marking is at -999 in amavis conf)[15:05:47] <wdp> wsxws, do you have @local_domains_maps defined with all domains you want to mark?[15:05:53] <wdp> (in amavisd.conf)[15:06:00] *** war has quit IRC[15:06:02] <wdp> if not, you should. ;)[15:06:07] <wsxws> a sec ill have a look at that[15:06:51] <wdp> i had such a problem, spamassassin wasn't rewriting the subject - but according to the log it was filtering with spamassassin. after i created such a map with all domains, it was working fine.[15:07:04] *** Lap_64 has quit IRC[15:07:06] *** jelly has quit IRC[15:07:08] *** hark has joined #postfix[15:07:29] *** PierreTramo has joined #postfix[15:07:37] *** jelly has joined #postfix[15:07:42] *** apeiron has joined #postfix[15:08:01] <wsxws> hm[15:08:08] <wsxws> how to add more than one domain there ?[15:08:20] <wsxws> and do i really need to add all of them one by ome ?[15:08:33] <wdp> wsxws, @local_domains_maps = ( [ ".$mydomain.", 'example.com', 'foobar.tld' ] );[15:08:34] <wsxws> (this machine here has 6 ips and plenty of domains)[15:08:45] <wsxws> ah[15:08:50] *** jelly has quit IRC[15:09:11] <wsxws> those '' are needed too ?[15:09:13] <wdp> wsxws, try google if there are better ways. i do it this way.[15:09:22] *** jelly has joined #postfix[15:09:42] <wdp> try it. amavisd will give errors if it's incorrect ;)[15:10:11] *** ming_zym has joined #postfix[15:11:13] <wdp> so someone here an idea cause of the relaying i mentioned[15:11:15] <wdp> ?[15:12:06] <wdp> wsxws, tell me if that was it ;)[15:12:14] <wsxws> @local_domains_acl = ([ ".$mydomain", 'wsxws.ws'] ); <<-- i have just thatone[15:12:25] <wsxws> there is no @local_domains_maps[15:12:50] <wdp> *shrugs*[15:12:55] <wdp> maybe version differences[15:13:22] *** jelly has quit IRC[15:13:27] <wdp> is it working now?[15:13:28] *** f3ew has joined #postfix[15:13:33] <wsxws> nope[15:13:44] <wdp> which version of amavisd you have?[15:13:54] *** jelly has joined #postfix[15:13:55] <wsxws> amavis-new[15:14:00] <wdp> version..[15:14:10] <Dominian> er yeah[15:14:11] <wdp> (thats the thing with the numbers.. )[15:15:00] <wsxws> i know waht version mean, but i dunno how to check[15:15:09] <wdp> which distribution you use?[15:15:13] <wsxws> ut its installed yesterday[15:15:18] <wsxws> debian etch[15:15:27] *** felix_da_catz has quit IRC[15:15:33] <wdp> uh, i used debian last time ago 5 years[15:15:50] <wdp> debian has a package manager... that should give you the version..[15:15:53] <wsxws> i installed the whole bunch, added the clamav volatile project and did and apt-get upgrade[15:16:01] <wsxws> so all versions should be uptodate[15:16:11] <wdp> wsxws, you could try renaming local_domains_acl to local_domains_maps[15:16:20] <wsxws> sure, a sec[15:16:21] <wdp> if thats not working, i have no idea.[15:16:22] *** Edward321 has joined #postfix[15:17:24] *** Edward123 has quit IRC[15:17:33] *** Edward321 is now known as Edward123[15:18:19] <wsxws> hm[15:19:28] <wsxws> i have x-sieve, x-virus, x-authenticated, x-provags and x-y but still ni x-spam[15:19:36] <wsxws> ni = no[15:19:56] *** Joelwork has quit IRC[15:20:04] <wsxws> like i had before the changes as well[15:20:09] <wsxws> means no changes[15:20:22] <wdp> so, dunno, sorry.[15:20:39] *** Azio is now known as dembug[15:20:51] *** _Scott-Mc has quit IRC[15:21:05] *** _Scott-Mc has joined #postfix[15:21:50] <wsxws> amavisd-new_1%3a2.4.2-6.1_all.deb[15:21:59] *** CrummyGummy has quit IRC[15:24:53] <wsxws> okies, mill give up 4 today[15:24:55] <wsxws> cu all[15:28:38] *** Zathraz has joined #postfix[15:34:17] *** |DeuS| has joined #postfix[15:36:58] *** wsxws has quit IRC[15:37:49] *** aozturk has quit IRC[15:38:18] *** kernel_ has quit IRC[15:39:41] *** GutterPunk has joined #postfix[15:40:52] *** eikke has joined #postfix[15:43:30] *** elvedin has joined #postfix[15:46:33] *** githogori has quit IRC[15:57:22] *** hal1on has joined #postfix[16:05:31] *** Sburk has joined #postfix[16:09:29] *** felix_da_catz has joined #postfix[16:09:41] *** hparker has joined #postfix[16:09:52] *** solar_ant has quit IRC[16:10:46] *** DeuS__ has joined #postfix[16:12:57] *** |DeuS| has quit IRC[16:13:58] <Sburk> Hello experts and good morning. I have a postfix server setup as an incoming SMTP server. It basically filters the email for spam and then sends it to my exchange server. Well, last night the exchange server went down. All incoming email was rejected. Shouldn't incoming email be queing up in postfix for when exchange comes back online?[16:14:34] <Sburk> I know that incoming mail is going through postfix prior to exchange. I checked that[16:16:48] <hparker> Depends on how you're doing recipient validation[16:17:30] <hparker> Unless you're exporting a list and sending it to the postfix server, it wouldn't know who's valid[16:17:50] <jpalmer> and chances are the rejections were soft bounces, meaning the remote servers will just re-send later[16:18:19] <hparker> You just have to wait a bit for your morning pr0n via email[16:18:21] <Sburk> ahhh. It seems as though some of our users did not receive email from last night[16:18:29] <Sburk> true[16:18:44] <jpalmer> (if it didn't softbounce, you need to revisit your implementation.)[16:19:20] <rob0> Yeah, lacking the LOGS of these rejections, we're just guessing.[16:19:26] <Sburk> Yeah, I think I will need to do so. This was not implemented by me. Is there some documentation for this sort of thing that you would recommend?[16:19:45] <hparker> postfix.org is a good starting point[16:20:14] <jpalmer> you can also cache recipient validation on the postfix machine, so that "already verified" recipients would have the mail accepted and queued by postfix in the future.[16:20:44] <rob0> You'll find about a million clueless MSexChange admins asking the same question over the years, on the postfix-users list, and getting a few clueful answers.[16:21:25] <Sburk> hmmm[16:21:58] <jpalmer> Sburk: you'd probably want to do that recipient validation cache anyway, to save resources on both the postfix and exchange machines.[16:22:26] <Sburk> being as that I did not implement postfix....I was unfamiliar with this behavior. rob0 if you don't care to be involved in discussion then....simply dont.[16:23:18] <jpalmer> Sburk: his comment was indeed accurate. in your search, you are going to find TONS of questions, and only a few clued answers. no need to bite the hands that feed.[16:24:06] <Sburk> oh woops. I'm sorry. I completely misread your comments rob0[16:24:13] <Sburk> I take that all back[16:25:07] <Sburk> I quickly glimpsed over the txt while multi-tasking and thought you were flaming me. Deepest appologies. Thx for the help guys[16:25:33] <Sburk> because I am in fact a clueless msexchange admin when it comes to postfix ;)[16:25:53] *** eikke has quit IRC[16:26:24] <Sburk> jpalmer: I already have the recipients caching in place. Thx![16:26:56] *** Lap_64 has joined #postfix[16:27:01] <jpalmer> Sburk: if so, why was postfix rejecting cached recipients. surely the majority of your users receive mail often enough to remain in that cache?[16:27:40] <Sburk> correct. I should reword that. I THINK I already have that in place.[16:28:33] <Sburk> it'a almost like postfix is not holding mail at all when exchange goes down[16:28:44] <Zathraz> Hi, I have a postfix machine (Debian Etch) which should become the new mailserver. Direct linking it to smtp.xs4all.nl does not work. Timeout issues, defered mail, etc. When I forward the mail to the old Exim server in the same LAN it is forwarded to smtp.xs4all.nl Any idea where to look?[16:28:51] <Sburk> I'm stumbling through to try and find you some logs[16:29:36] *** brancaleone has quit IRC[16:29:39] *** Nockian has quit IRC[16:29:44] <jpalmer> that sounds to me like no cache in place, and postfix is authenticating valid users on an as-needed basis from exchange. which is fine IF it's softbouncing the mails when exchange is unavailable.[16:30:12] <Sburk> roger that jpalmer. Thank you! I will research soft bouncing and cache[16:30:54] <jpalmer> if it's softbouncing those mails, then the users will get them later.[16:31:40] <jpalmer> and just a caveat. I'm not talking about the global softbounce setting in postfix. you absolutely don't want that on a live production machine unless you are making changes to the config.[16:32:13] <Sburk> good info. thx palmer[16:36:50] *** eikke has joined #postfix[16:38:59] *** brancaleone has joined #postfix[16:40:36] *** Danielss89 has left #postfix[16:42:01] *** mcf3782 has joined #postfix[16:45:56] <mcf3782> Anyone in here have one of the Yellow Machine NAS boxes from the now defunct Anthology Solutions? It appears to use postfix for sending mail, but I'm not having much luck. Would love to talk to someone who's gotten it to work. Baring that, the error I'm fighting is " 550 [PERMFAIL] bellsouth.net requires valid sender domain)".[16:47:10] <LoRez> sounds like your box is using a domain that doesn't exist.[16:47:14] <mcf3782> doesn't matter what domain I try to send mail to (bellsouth.net, gmail.com, mindspring.com...) the error is always the same "...requires valid sender domain)[16:48:21] <LoRez> it's where you're sending it from that's important[16:48:25] <rob0> Normally the sender address is set in the MUA. How are you sending this mail?[16:49:03] <mcf3782> mail -v montef at gmail dot com[16:50:36] *** Lap_64 has quit IRC[16:51:13] <mcf3782> looking at the messages that get logged in /var/log/mail.log, it doesn't appear that the outbound message that gets handed to gmail.com has a valid 'from' field:[16:51:28] <rob0> ok, that's a rather primitive MUA. Some incarnations thereof do have the ability to set a sender address, try "man mail". Also ...[16:51:34] <rob0> !myorigin[16:51:34] <knoba> rob0: "myorigin" : a configuration parameter in the main.cf: The default domain name that locally-posted mail appears to come from, and that locally posted mail is delivered to. The default $myhostname, which is fine for small sites. If you run a domain with multiple machines, you should (1) change this to $mydomain and (2) set up a domain-wide alias database that aliases each user to user at that dot users.mailhost.[16:56:23] <mcf3782> so the default in /etc/postfix/main.cf of "myorigin=$myhostname" isn't correct? "myhostname=YM900CEF.bellsouth.net". If I read that right, then myorigin and myhostname should both eval to "YM900CEF.bellsouth.net", which is the hostname of the machine.[16:57:54] *** sepski has joined #postfix[17:01:54] <rob0> "The default ... isn't correct?" Huh? Wietse picks a default which is most likely to DTRT in the largest number of deployments. Postfix defaults in no way relieve the admin of the responsibility to determine and to set the correct settings at each instance.[17:03:27] * cpm defaults to rob0[17:04:02] <rob0> check_cpm_access static:reject[17:04:09] * cpm outs[17:04:12] <cpm> pouts even[17:04:18] <cpm> Doh![17:04:33] * cpm waits for it[17:06:07] <hparker> puts out?[17:06:36] <rob0> omg tmi[17:11:33] <jpalmer> I'm impressed. almost 24 hours now without a sporking[17:11:58] * cpm digs around in his desk, , , can't find a spork[17:12:32] * jpalmer picked them all up and melted them. even the fancy old metal ones.[17:12:39] <mcf3782> Where does this "(2) set up a domain-wide alias database that aliases each user" get done?[17:12:58] <jpalmer> now, I use the molten mess in a game I made up. 'sporkball'[17:13:17] *** j_s has joined #postfix[17:13:18] <cpm> a poll, when was the first time you encountered a spork, and what did it look like?[17:14:13] <mcf3782> funny.. I went to a Halloween party last weekend. One of the activities was a scavenger hunt. One of the items we had to find was a spork. Surprisingly difficult to get here in Atlanta, GA.[17:14:15] <jpalmer> cpm: probably in gradeschool. I hated it on pizza day. always broke. thats when we decided it was cooler to roll your square pizza up, and eat it like a burrito.[17:14:27] <hparker> mcf3782: User definable, but usually /etc/aliases, /etc/mail/aliases, /etc/postfix/aliases[17:14:29] <cpm> heh[17:14:48] <hparker> mcf3782: KFC usually has them[17:15:03] <cpm> first spork I encountered was as a child in the 60s, buddy had a pocket knife with a spork on it.[17:15:16] *** war has joined #postfix[17:15:17] <mcf3782> nope. KFC was the first place we tried.[17:15:26] <mcf3782> Finally found them at Taco Bell[17:15:47] <cpm> heh[17:15:50] * hparker doesn't remember the first spork he saw.. Also doesn't remember the 60s nor most of the 70s, and not a lot of the 80s[17:16:15] <hparker> mcf3782: ouch.. So then just what is KFC good for if not to get sporks?[17:16:16] <mcf3782> too much LDS?[17:16:17] <mcf3782> ;)[17:16:22] * jpalmer wasn't alive for the 60's, and only partially for the 70's. for once, I don't feel like an old man on this network.[17:16:30] <hparker> mcf3782: That was the 80s, yes[17:16:40] <mcf3782> Around here, KFC is not good for much of anything.[17:17:22] * cpm napped on the bottle under Sputnik skies[17:17:42] * hparker bets that was uncomfortable[17:18:07] <mcf3782> It looks like /etc/aliases on this box is fully populated.[17:22:27] *** jelly has quit IRC[17:22:58] *** jelly has joined #postfix[17:26:59] *** cilly has quit IRC[17:30:04] *** theblackbox has joined #postfix[17:30:56] <rob0> cpm, were you the one on the grassy knoll in Dallas?[17:31:03] *** tmjb has quit IRC[17:31:22] <hparker> lol[17:31:37] <jelly> a-ha! How do _you_ know there was only one person?[17:32:01] * cpm isn't paying attention[17:32:36] *** jelly has quit IRC[17:32:49] <cpm> see?[17:32:53] <cpm> that'll learn'n[17:32:54] <cpm> m[17:33:08] *** jelly has joined #postfix[17:34:28] *** phnord has quit IRC[17:37:07] *** Tjikkun_work has quit IRC[17:38:05] <Zathraz> Hi folks. I am using Postfix in an OMA (openmailadmin) environment (mysql, SASL). Server is behind a firewall. I prefer not to install anything on the firewall machine or reconfig it. Can I reconfig just Postfix or should I also do something on the fw machine too (hope not)?[17:40:06] <hparker> you'll need to forward at least port 25 to it[17:42:15] * cpm forwards hparker[17:42:27] <Zathraz> hmm. Troublesome. I want to test this server (it should both send and receive mail) but in the meantime the old exim server should still do it's job[17:42:32] * hparker moves along[17:42:35] <Zathraz> lol[17:43:26] <hparker> that might get kinda hard if it's forwarded through the same firewall as they need the same port[17:43:40] <mcf3782> Hey.. if forwarding the entire user really works; then I have a whole list! Do I get to pick where they get forwarded to?! :-D[17:43:41] <Zathraz> yep[17:44:07] <Zathraz> can I request for something like the Bahama's?[17:44:46] <mcf3782> mv Zathraz /dev/Bahamas[17:45:17] <Zathraz> currently I get a timeout error from the ISPs SMTP server. That is caused by the remote SMTP server trying to connect back to Postfix?[17:45:55] <hparker> Might be if they're doing SAV or something... Are you using an email address that exists on the exim system?[17:48:07] *** veepster has quit IRC[17:52:24] <Zathraz> as a sender I currently use 'root'[17:52:36] <Zathraz> which exists ofcourse on the exim server ;-)[17:52:50] <hparker> What's the reject message?[17:53:09] <mcf3782> mv $ME $LUNCH[17:53:14] *** mcf3782 has left #postfix[17:53:28] <Zathraz> connect to smtp.xs4all.nl[194.109.6.51]: Connection timed out (port 25)[17:53:49] <Zathraz> followed by: relay=none, delay=30, delays=0.04/0/30/0, dsn=4.4.1, status=deferred (connect to smtp.xs4all.nl[194.109.6.51]: Connection timed out)[17:54:03] <hparker> that sounds like a firewall... telnet 192.109.6.51 25[17:54:06] <Zathraz> which was appended after the mailID and emailadres[17:54:54] *** ming_zym has quit IRC[17:55:34] <Zathraz> both servers do not seem to be too keen to connect to that server[17:55:47] <Zathraz> using telnet that is[17:55:57] <cpm> telnet is the base.[17:56:07] <Zathraz> yep. I know[17:56:22] <Zathraz> but on one machine mail is send through that host regardless[17:57:13] *** bkw has joined #postfix[17:58:01] *** Edward123 has quit IRC[18:01:17] <Zathraz> k. I copied your telnet command. However there was a typo in the IP[18:01:36] <Zathraz> so now on the exim machine telnet works, on the other machine it doesn't[18:01:43] <Zathraz> sounds logical so far[18:05:16] <Zathraz> to make things a lot more wierd: I can access that machine (the postfix server) from the LAN without problems. But I cannot connect to it through a VPN connection. All other servers can be reached. Wierd[18:05:37] *** apeiron has quit IRC[18:06:38] <hparker> sounds like an anal firewall I'd setup[18:07:47] <Zathraz> I didn't set it up. has been running for years without problems until now I suppose :-([18:08:40] <hparker> If you don't need to get to the IP, you can't... So then when adding/testing you have to alter the firewall[18:10:13] *** poisdeux has quit IRC[18:10:46] * cpm recommends taking a firewall down when doing this stuff, otherwise you chase too many problems, that aren't problems.[18:10:55] <Zathraz> true[18:11:12] <Zathraz> yet, with fw down I the issue seems the same[18:11:39] *** apeiron has joined #postfix[18:18:39] <Zathraz> hmmm. Made a very strange error. Wierd I never spotted it before. Appologies. Something was foobar in the IP routing on the postfix server[18:25:13] *** apeiron has quit IRC[18:26:54] <Zathraz> everything is now working as it should[18:27:02] <Zathraz> thanks a lot guys[18:27:12] <Zathraz> got to go now. bye[18:27:19] *** Zathraz has left #postfix[18:32:30] *** r0sk has left #postfix[18:35:31] *** githogori has joined #postfix[18:36:03] *** solar_ant has joined #postfix[18:36:34] *** jelly has quit IRC[18:40:38] *** jelly has joined #postfix[18:47:53] *** cilly has joined #postfix[18:54:34] *** jelly has quit IRC[18:54:43] *** mark-use has joined #postfix[18:55:16] *** jelly has joined #postfix[19:06:48] *** weggpod has joined #postfix[19:07:53] *** havvg has quit IRC[19:13:24] *** amrit|zzz is now known as amrit|wrk[19:18:39] *** brancaleone has quit IRC[19:20:43] *** _Scott-Mc is now known as Scott-Mc[19:25:48] <wdp> is it possible to have relay_domains point to transport maps, so that i don't need to define domains twice?[19:26:59] <rob0> sure, but be careful. If you ever need a special transport for some other domain, you might want to add a second lookup to transport_maps.[19:31:39] <sysmonk> wdp: damn it, when i see your nick, i think of my co-worker![19:31:54] <wdp> i think i don't understand. i thought simply to use transport maps = hash:/somefile and then relay_domains = $transport_maps[19:34:23] <sysmonk> wdp: rob0 says, that you may be needing some special transport for some domain, which shouldn't be in relay_domains[19:34:38] <sysmonk> for that, you'd have to specify the hash:/file for both transport_maps and relay_domains[19:34:42] <sysmonk> and not relay_domains = $transport_maps[19:35:00] <sysmonk> i.e. you'd have relay_domains = hash:/file1; transport_maps = hash:/file1[19:35:14] <wdp> isn't that the same?[19:35:19] <sysmonk> yes, same[19:35:20] <sysmonk> buuuut[19:35:31] <sysmonk> if you will do this like relay_domains = $transport_maps[19:35:43] <sysmonk> and, in the future you'd want to have a special transport for some domain[19:35:52] <wdp> whats a special transport for example?[19:35:53] <sysmonk> you won't be able to add it in transport_maps[19:36:14] <wdp> in the transport_maps file i have domain.tld smtp:[some ip][19:36:15] <sysmonk> wdp: uh... i.e., what if you'd want to add a domain, which should be only in transport_maps[19:36:18] <sysmonk> and not in relay_domains ?[19:36:36] <wdp> i think, this will not happen[19:36:48] <sysmonk> in your situation ( with relay_domains = $transport_maps ) - it's not possible[19:36:51] <wdp> but ty, i understand now[19:37:07] <sysmonk> the correct solution would be relay_domains = hash:/file1[19:37:09] <wdp> could someone here send me a testmail? ^^[19:37:17] <sysmonk> and transport_maps = hash:/file1, hash:/file2[19:37:51] <sysmonk> wdp: sure :)[19:38:19] *** marc7 has joined #postfix[19:38:21] * sysmonk redirects all spam to wdp email addie, muahahahaha!!! :)[19:38:48] <wdp> ^^[19:39:06] <sysmonk> sent[19:39:21] <sysmonk> greylisting in action[19:39:49] <wdp> whats .lt?[19:39:54] <sysmonk> lithuania[19:40:17] *** weggpod has quit IRC[19:40:30] <wdp> i get lots of spam from there (filtered by our gateway)[19:40:35] <sysmonk> really ?[19:40:46] <sysmonk> i think you get spam for .it[19:40:47] <sysmonk> not .lt[19:40:57] <wdp> helo=<78-61-232-40.ip.zebra.lt>[19:41:04] <sysmonk> ah, those assholes :P[19:41:38] <sysmonk> zebra.lt is a shitty isp, with lot's of lame users :)[19:41:54] <sysmonk> there's also one of zebra.lt sysadmins in this channel :PPP[19:42:08] <sysmonk> (not currently online)[19:42:31] <wdp> *shrugs*[19:42:40] <wdp> it's blocked using blacklists.[19:45:05] <wdp> i see sometimes using @domain.tld .domain.tld and simply domain.tld - .domain.tld means inclusive subdomains[19:45:08] <wdp> whats @domain.tld?[19:46:01] <sysmonk> ?[19:46:09] <sysmonk> tld = top level domain[19:46:17] <wdp> i mean in hash or access files[19:46:43] <sysmonk> depends where ;)[19:46:50] <wdp> transportmaps[19:47:34] <sysmonk> never saw a @domain.tld in transport_maps, always used domain.tld or user at domain dot tld[19:48:26] *** etaylor has joined #postfix[19:49:25] *** Xzisted has joined #postfix[19:49:29] *** bkw has quit IRC[19:49:40] <Xzisted> hey all...i have a problem i need help with[19:50:06] <wdp> good to know[19:50:12] <wdp> would be better to know your problem..[19:50:57] * sysmonk stands up and starts talking: Hi! I'm alex! I'm an alcoholic.[19:51:33] <wdp> and we're still waiting for the problem description[19:51:39] <Xzisted> i am setting up a mail server and i need it to do the following: mail from example.com users on trusted networks (say to user at gmail dot com) gets sent straight to gmail, however mail FOR example.com users from gmail (mailto: user at example dot com mailfrom:user at gmail dot com) comes in, gets processed, then forwarded to a relayhost[19:52:02] <Xzisted> that relayhost being exchange[19:52:21] <Xzisted> i set the relayhost parameter, however, its taking mail for gmail users and trying to pass it to the relayhost[19:52:31] <Xzisted> which is bad[19:53:16] * sysmonk didn't understand the situation[19:53:17] *** jelly has quit IRC[19:54:57] *** chris_ has quit IRC[19:55:01] *** jelly has joined #postfix[19:55:10] <Xzisted> sysmonk: lets say you work at example.com. you are at your desk, you send an email to jarjar at gmail dot com. that mail goes to the postfix server and the postfix server should send it straight to gmail. jarjar at gmail dot com replies to you. it hits the postfix server, but instead of delivering it to a local mailbox, it needs to forward it to another internal server before it can be delivered to you.[19:55:41] <sysmonk> Xzisted: transport_maps[19:55:54] <sysmonk> !transport_maps[19:55:55] <knoba> sysmonk: "transport_maps" : a configuration parameter in the main.cf: Optional lookup tables with mappings from recipient address to (message delivery transport, next-hop destination). See transport(5) for details.[19:58:04] *** Twinkletoes has quit IRC[19:59:33] <Xzisted> sysmonk: that looks like what i need. thanks[20:02:04] <wdp> if i have defined relay domains, transport maps and relay recipients maps, and i have many domains in relay domains and transport maps. in relay domains is for example.net - i have no address with example.net in relay recipient maps, what will happen? that will let all mail for domain example.net go through and the mailserver is a source for backscatter mail, correct?[20:02:09] *** collink_ has joined #postfix[20:02:45] *** twister1207 has joined #postfix[20:03:11] <twister1207> hello... are you talking german or in english ?[20:03:16] <collink_> All of the sudden, my postfix setup doesn't want to let me delete messages. I'm getting this error: "The IMAP command “UID COPY” (to Deleted Messages) failed for the mailbox “INBOX” with server error: Error in IMAP command received by server.."[20:03:18] *** marc7 has quit IRC[20:03:28] <wdp> twister1207, i'm german. we're talking english[20:03:41] <twister1207> I'm german,too...[20:04:36] <collink_> Has anyone ever seen this?[20:04:47] <sysmonk> collink_: that's not postfix problem, it's imap problem[20:04:57] <collink_> oh....well do you have any idea?[20:04:58] <sysmonk> postfix is a SMTP server, not IMAP server[20:05:09] <twister1207> Nov 2 19:48:30 SERVER postfix/qmgr[7440]: 3D6F052469E: to=<register at harbich-vhm dot de>, relay=none, delay=2422, status=deferred (delivery temporarily suspended: Host or domain not found. Name service error for name=smtp.1und1.de type=MX: Host not found, try again)[20:05:24] <twister1207> what is wrong here ?[20:06:38] <collink_> "Host or domain not found."[20:07:56] <sysmonk> i just don't understand how did it expand harbich-vhm.de to smtp.1und1.de[20:08:05] <sysmonk> twister1207: using "smart relay" ? ( relayhost )[20:08:08] <twister1207> twister1207@SERVER:~$ dig mx smtp.1und1.de[20:08:10] <twister1207> ; <<>> DiG 9.3.2 <<>> mx smtp.1und1.de[20:08:11] <twister1207> ;; global options: printcmd[20:08:15] <twister1207> ;; Got answer:[20:08:17] <twister1207> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 48973[20:08:20] <twister1207> ;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0[20:08:21] <twister1207> ;; QUESTION SECTION:[20:08:23] <sysmonk> twister1207: use pastebin for pasting[20:08:24] <twister1207> ;smtp.1und1.de. IN MX[20:08:25] <twister1207> ;; AUTHORITY SECTION:[20:08:28] <twister1207> 1und1.de. 3042 IN SOA nsa.schlund.de. hostmaster.schlund.de. 2007110206 10800 3600 604800 86400[20:08:31] <twister1207> ;; Query time: 49 msec[20:08:32] <twister1207> ;; SERVER: 192.168.178.1#53(192.168.178.1)[20:08:35] <twister1207> ;; WHEN: Fri Nov 2 20:07:53 2007[20:08:36] <twister1207> ;; MSG SIZE rcvd: 90[20:10:06] <wdp> twister1207, sorry, i can't give private support in private messages, it's not a problem with you, the thing is, i have no time. If you need an administrator, you should hire someone.[20:10:31] * sysmonk puts his hand up[20:10:50] <wdp> twister1207, and most things "how to get a smart relay working" are documentated. just read on at postfix.org[20:10:53] <sysmonk> do i smell money? :)[20:11:08] * cpm smells monkey[20:11:11] <cpm> ewww[20:11:23] <wdp> lo,l[20:11:39] <sysmonk> cpm: that was a used toilet paper, dumbass! :)[20:11:49] <cpm> ewwwww!~[20:12:02] * cpm needs to go wash up now[20:12:08] <cpm> Have a good weekend folks[20:12:13] *** cpm has quit IRC[20:12:34] <sysmonk> so, where does that smell of money come from ? :)[20:12:40] <wdp> i don't understand this backscatter thingy[20:12:50] *** weggpod has joined #postfix[20:13:05] <sysmonk> let me say it in other words[20:13:14] <sysmonk> where does the smell of "Geld" come from? :)[20:13:45] <wdp> i don't like german[20:13:54] <wdp> :p[20:14:21] <sysmonk> uh, i don't know german ;P[20:15:22] <twister1207> sysmonk : i don't use smart reley[20:15:30] <twister1207> relay[20:18:22] <twister1207> Nov 2 20:13:53 SERVER postfix/smtp[7627]: CDE7852468D: to=<register at harbich-vhm dot de>, relay=smtp.1und1.de[212.227.15.183], delay=1, status==deferred (host smtp.1und1.de[212.227.15.183] said: 421 dns lookup failed for sender domain (in reply to MAIL FROM command))[20:18:53] <sysmonk> it reject the mail because of sender domain[20:19:00] <sysmonk> which we don't see in your pasted log line[20:19:07] <sysmonk> also, please, consider using a PASTEBIN[20:19:33] <twister1207> should i post the main.cn or the mail.log ?[20:20:09] <rob0> NO[20:20:30] <twister1207> another file ?[20:20:47] <rob0> "postconf -n" and a selection of the log, not the whole thing.[20:21:13] <rob0> Logs covering at least one whole message.[20:21:47] <rob0> and if you do another multi-line paste here, /ignore[20:21:59] <twister1207> http://ubuntuusers.de/paste/17390/[20:22:35] <sysmonk> huh[20:22:36] <sysmonk> relayhost = smtp.1und1.de:25[20:22:40] <sysmonk> 11-02 21:15:23 < twister1207> sysmonk : i don't use smart reley[20:22:47] <sysmonk> yeah, whatever...[20:23:12] <jelly> maybe it's not that smart.[20:24:02] *** emo_ninja has joined #postfix[20:24:03] <sysmonk> twister1207: sorry, do you use stupid relay?[20:24:04] <sysmonk> ;P[20:24:25] <twister1207> i don't know what a stupid reley is...[20:24:29] <emo_ninja> is there a way to disable forwarding to aliases after a content filter is done[20:24:32] <sysmonk> twister1207: pastebin a grep CDE7852468D /var/log/maillog or whatever is the mail log file on your system[20:24:39] <emo_ninja> or before[20:25:28] <wdp> sysmonk, can you test mail me again please? ;)[20:25:29] <wdp> +send[20:25:32] <wdp> +a[20:25:34] <wdp> +to[20:25:49] <emo_ninja> I'm getting duplicates because when mail comes in for user X it's aliased to go to user X, User Y => content filter (clamd) => back to another postfix port => aliased again,[20:25:54] <twister1207> http://ubuntuusers.de/paste/17391/[20:25:58] <emo_ninja> so I'm duplicating the aliases[20:26:59] <sysmonk> sent again wdp[20:27:41] <wdp> sysmonk, funny, working, ty for your help. maybe i can be helpful someday too ;)[20:28:06] <sysmonk> wdp: you'd have to check if it works with unknown recipients, not with known :PP[20:28:23] <wdp> hm[20:28:35] <sysmonk> that's the main reason of using that stuff :P[20:29:44] <sysmonk> also, you could look at using address_verify_map[20:29:52] <sysmonk> if you need it, ofcorse :)[20:30:01] *** hemry has joined #postfix[20:31:01] <wdp> cool[20:31:04] <wdp> yeah, it's working.[20:31:14] <twister1207> sysmonk : http://ubuntuusers.de/paste/17391/[20:31:23] <wdp> sysmonk, that gives the answer from another mailserver "Recipient address rejected: User unknown in virtual mailbox table"[20:31:30] <emo_ninja> ah perhaps: receive_override_options=no_address_mappings in the process that sends it to a content filter[20:31:31] <wdp> sysmonk, ty.[20:32:12] <wdp> sysmonk, and this will not generate mailer-daemons?[20:32:13] <sysmonk> twister1207: change myhostname in main.cf[20:32:33] <sysmonk> wdp: it will be rejected in the smtpd_recipient_restrictions[20:32:45] <sysmonk> so, it will be rejected BEFORE the mail will be queued[20:32:47] <twister1207> sysmonk : to ?[20:32:56] <sysmonk> twister1207: fqdn of your server?[20:32:57] <wdp> ty.[20:33:33] <twister1207> sysmonk : to the name from my server ?[20:33:44] <sysmonk> twister1207: uh, what's the ip of your server?[20:35:41] <sysmonk> it should be something like yourservername.yourdomain.yourtld[20:35:41] <sysmonk> i.e. something.twister1207.de[20:35:41] <sysmonk> or a rDNS of your server[20:35:41] <sysmonk> anything which points to your server[20:35:41] <sysmonk> currently it's just "SERVER"[20:35:42] <twister1207> 192.168.178.23[20:35:43] <collink_> INTERNAL![20:35:43] <collink_> NOOOOOOOOOO![20:35:43] <collink_> I know[20:35:44] <collink_> sry...last one was a mischan[20:35:47] *** Mattias has joined #postfix[20:36:02] <sysmonk> twister1207: uh, nevermind... any domains that point to your server?[20:36:06] <wdp> lol[20:36:17] <rob0> $myhostname should be a FQDN which resolves to your [external of course] IP address[20:36:35] <sysmonk> rob0: thanks for saying that in one line :P[20:36:46] <sysmonk> rob0: maybe translation to german will help ? :P[20:36:55] <rob0> !myorigin[20:36:56] <knoba> rob0: "myorigin" : a configuration parameter in the main.cf: The default domain name that locally-posted mail appears to come from, and that locally posted mail is delivered to. The default $myhostname, which is fine for small sites. If you run a domain with multiple machines, you should (1) change this to $mydomain and (2) set up a domain-wide alias database that aliases each user to user at that dot users.mailhost.[20:37:22] <wdp> twister1207, $myhostname muss eine externe ip sein[20:37:41] <twister1207> wdp : also zu meinem home server,richtig ?[20:37:44] <twister1207> oder zur domain ?[20:37:54] <wdp> twister1207, deine internet ip[20:38:07] <twister1207> wdp : zu mienem server,richtig ?[20:38:23] *** havvg has joined #postfix[20:38:28] * sysmonk knows some words too! shveine or something :P[20:38:48] <wdp> twister1207, der Mailserver hat eine lokale ip. Und wie ist die EXTERNE ip ueber die der Mailserver angesprochen werden kann?[20:39:04] <twister1207> bekommt man dei auch über die console raus ?[20:39:07] * sysmonk feels stupid reading all this stuff[20:39:10] <wdp> ja. "ifconfig ppp0"[20:39:26] <wdp> wenn das bei dir zuhause ist[20:39:27] <sysmonk> uh, maybe you'd want to write that in your PM's? :P[20:39:32] <wdp> no...[20:39:35] *** meandtheshell has joined #postfix[20:39:36] <wdp> sorry ;)[20:40:13] * hparker looks for his sekrit decoder glasses[20:41:18] <twister1207> 84.173.32.228[20:41:56] <rob0> 228.32.173.84.in-addr.arpa. 86400 IN PTR p54AD20E4.dip0.t-ipconnect.de.[20:42:09] <sysmonk> everybody ping -f 84.173.32.228 now![20:42:10] <rob0> is that a dialup IP?[20:42:21] <sysmonk> rob0: uh, telling by ppp0 - yes :P[20:42:22] *** weggpod has quit IRC[20:42:41] <rob0> could be ADSL[20:42:50] <sysmonk> rob0: which is dialup also :)[20:43:26] <rob0> well, usually a bit faster than a modem :)[20:43:59] <sysmonk> a liiiitle bit faster :P[20:44:23] <rob0> Anyway, you can't expect to run a MTA from a dynamic IP address, regardless of type of connection.[20:44:29] <wdp> rob0, yeah.[20:44:35] <wdp> rob0, that is a german dsl connection.[20:45:02] <sysmonk> rob0: but a work-around would be to have those noip.org hosts[20:45:06] <sysmonk> or whatever[20:46:06] <twister1207> it don't work ? is that right ?[20:46:24] <wdp> reject_unverified_recipient[20:46:33] <wdp> will that skip further restrictions?[20:46:56] <wdp> twister1207, ja. es wird nicht gehen. Normalerweise installiert man keinen Mailserver auf einem rechner "zuhause" mit einer dynamischen ip - wenn du damit ins internet willst.[20:47:04] <wdp> twister1207, dafuer jibbet server mit statischen ip's.[20:47:33] <twister1207> der dient eigentlich nur zurm datenaustausch mit sdamba im lokalen netz[20:47:46] <twister1207> internet is damit nciht drin.. außer für mails abholen dachte ich...[20:47:53] <twister1207> deswegen auch postfix...[20:48:06] <wdp> twister1207, erklaers auf englisch, dann koennen die anderen helfen und verstehen auch was du da machst.[20:48:10] <sysmonk> wdp: uh, what's your unverified_recipient_reject_code ?[20:48:11] *** havvg has quit IRC[20:48:27] <wdp> sysmonk, haven't set.[20:48:45] <sysmonk> 450 then[20:48:51] <sysmonk> isn't 450 a pemanent error ?[20:48:54] <wdp> sysmonk, i have some "status=deliverable" in log. thats why i ask - i hope that this rule isn't given an "OK" and further rules are skipped.[20:49:11] <wdp> 554 is permanent error iirc.[20:49:28] <sysmonk> wdp: deliverable isn't the same as delivered[20:49:38] <wdp> ok.[20:49:39] <sysmonk> it just says that it CAN be delivered[20:49:44] <wdp> k ;)[20:49:50] <wdp> just wanted to make sure.[20:49:57] <sysmonk> it's the effect of reject_unverified_recipient[20:50:14] <shasta> sysmonk, 5xx are permanent, 4xx are temporary[20:50:22] <sysmonk> ah, damn :)[20:50:32] * sysmonk always has those two wrong :P[20:50:45] <rob0> Whatever the choice, $myhostname *must* resolve to the external IP, and you *must* use a real resolvable sender domain if you want anyone to accept your mail.[20:50:46] <sysmonk> wdp: you could also set the reject code to 550 then[20:51:12] *** lala has joined #postfix[20:51:16] <wdp> sysmonk, wouldn't the reject be cached or something like that? or is it always verifying then?[20:51:26] <lala> hello[20:51:34] <lala> i can help in postfix[20:51:43] <shasta> sysmonk, just remember that the bigger the first number is, the worse the error[20:51:45] <sysmonk> wdp: the reject is catched, look at the address_verify_map[20:51:56] <sysmonk> shasta: sure, like 200 ?:PP[20:51:58] <shasta> sysmonk, 2xx are "OK", 4xx are "maybe OK, but not now", 5xx are "no way"[20:52:14] <sysmonk> shasta: it was a joke :)[20:53:01] <sysmonk> i'm just always having problems remembering 2 things about smtp: the 4xx and 5xx, and the postfix delays=xx/xx/xx[20:53:07] <sysmonk> don't remember which is which :)[20:54:14] *** havvg has joined #postfix[20:54:51] <wdp> delay_warning_time (default: 0h)[20:54:54] <wdp> for what is that thing good?[20:55:22] <sysmonk> wdp: for dsn's[20:55:36] <sysmonk> if your client sends an email, you can set the delay_warning_time to ... i.e. 2 hours[20:55:55] <sysmonk> and if the email is still in your queue (for some reason) for longer than 2 hours - the clients gets a notification about it[20:56:06] <sysmonk> so that he knows, that the mail still isn't delivered[20:56:27] <wdp> sysmonk, so while i'm doing greylisting (5 minutes) i could set this to 5 minutes or?[20:57:10] <lala> i are could help me[20:57:21] <sysmonk> lala: i are baboon![20:57:32] <lala> baboon?[20:57:39] <wdp> sysmonk, than the sending mailserver would try to redeliver after 5 minutes, or not?[20:58:09] <sysmonk> wdp: uh, you've got it wrong :)[20:58:10] <collink_> lala: e was making fun of you grammar[20:58:16] <wdp> too bad ;)[20:58:29] <sysmonk> wdp: consider me being your smtp server client[20:58:31] <sysmonk> the outgoing one[20:58:39] <wdp> are you female?[20:58:50] <sysmonk> wdp: you set the delay_warning_time to 1hour[20:58:51] <wdp> (okay.. bad joke)[20:59:05] <sysmonk> now, i send an email to my friend, who uses some shitty mailserver[20:59:16] <sysmonk> and the destination mailserver is down[20:59:32] <sysmonk> so postfix has the mail in the queue, for 3 days ( default ) and tries to resend that mail[20:59:52] <sysmonk> after the mail in queue is 1 hour old, i ( teh client ) receive a NOTIFICATION, that the mail is STILL in the queue[20:59:59] <wdp> maximal_queue_lifetime = 4000s[21:00:05] <wdp> *giggle*[21:00:25] <sysmonk> k, the default is 5d now :) somehow i thought it's 3 days :)[21:00:26] <sysmonk> anyway[21:00:31] <wdp> no.[21:00:34] <wdp> i set it to 4000 seconds.[21:00:44] <sysmonk> wdp: i know, i just rechecked what's the default[21:00:54] <wdp> ic.[21:11:01] *** twister1207 has quit IRC[21:14:52] *** bureado has joined #postfix[21:18:36] <hparker> spork! http://farm3.static.flickr.com/2327/1697665718_c8f5be9d24.jpg[21:19:17] <bureado> Greetings. I'm currently using Postfix with an LDAP server. I'm using mail queries to dereference aliases from LDAP groups, and I have two 'mail' attribute entries for each user. How can I get only one of the two entries?[21:22:52] <sysmonk> bureado: size_limit ?[21:27:17] *** growltiger has quit IRC[21:32:00] *** b4 has joined #postfix[21:32:16] <bureado> sysmonk: Yeap, I've tried that, but I get a temporary error in the lookup.[21:32:20] *** hemry has quit IRC[21:33:10] *** hemry has joined #postfix[21:34:15] <sysmonk> bureado: ?[21:34:26] *** pirho has joined #postfix[21:34:37] <sysmonk> bureado: postmap -q gives a temporary error?[21:35:23] <sysmonk> pastebin the related cf file, maybe you've wrote something incorrectly ?[21:35:55] *** Zeit|awy has joined #postfix[21:38:12] *** mark-use has left #postfix[21:39:31] <bureado> sysmonk: Without restrictions -> http://pastebin.ca/759089[21:40:39] <sysmonk> uh, so it returns 2 results[21:40:54] <sysmonk> bureado: and how does that look in the ldap ?[21:40:57] <bureado> Yes, and I just need one. Otherwise mail will be dropped twice.[21:41:00] <sysmonk> mail = user@domaina[21:41:03] <sysmonk> mail = user@domainb[21:41:04] <bureado> sysmonk: Every user has two mail: attributes.[21:41:09] <sysmonk> or mail = user@domaina,user@domainb ?[21:41:16] <bureado> The former[21:41:59] <bureado> ldapsearch -x uid=user mail | grep '^mail:' | wc -l -> 2[21:43:02] <sysmonk> how about expansion_limit = 1 ?[21:43:21] <sysmonk> (i'm not using ldap so i'm just guessing around ...)[21:44:41] <bureado> postmap -q returns 'Expansion limit exceeded for key...' as a warning, but it doesn't display either[21:45:45] *** Tinozaure has quit IRC[21:46:10] <bureado> sysmonk: That's ok, I use LDAP but don't know about Postfix alias tables D:[21:47:09] *** Tinozaure has joined #postfix[21:49:03] <sysmonk> bureado: i'd play around with those params[21:49:33] <sysmonk> don't have how to replicate this, so can't help ;/[21:50:46] *** Zeit|awy has quit IRC[21:51:36] *** grigora has joined #postfix[21:52:46] <grigora> Hi, my server magically redirects mail sent to user1 at domain1 dot com to user2 at domain2 dot com but I cannot find where this is configured. I'd apprecaite any pointers ...[21:53:43] <bureado> sysmonk: It's ok, thank you very much indeed![21:56:28] *** Zeit|awy has joined #postfix[21:57:44] *** bureado has quit IRC[21:59:20] *** hemry has quit IRC[21:59:40] <grigora> i just pasted all the relevant config files - http://sial.org/pbot/28409[22:03:41] <grigora> anyone?[22:04:48] *** hemry has joined #postfix[22:04:59] <rob0> Munged and useless.[22:06:23] *** pitakill has joined #postfix[22:07:41] *** Mattias has quit IRC[22:08:11] <grigora> rob0: are you talking to me?[22:10:13] *** felix_da_catz has quit IRC[22:10:39] <rob0> domain1.com and domain2.com do not appear in your config, so you're stuck having to find a magical solution to the magical problem.[22:10:58] <grigora> rob0: actually i just figured out the issue[22:11:04] <grigora> rob0: never mind[22:12:39] <grigora> my other question is that when I comment out the smtpd_tls_* variables, I can no longer connect to the database to get the local_recipient_maps. Anyone knows why this is? considering that those settings are bogus.[22:16:20] <rob0> Hmm, I wouldn't see any connection there, even if the pgsql server was using TLS (which would be silly, as it's on localhost.)[22:16:28] <rob0> !postmapq[22:16:29] <knoba> rob0: "postmapq" : You can check your lookups with the postmap command. Example: if you defined "transport_maps = mysql:/etc/postfix/transport.cf" you may check this mapping by running "postmap -q domain.com mysql:/etc/postfix/transport.cf" and see if it works.[22:19:31] <grigora> rob0: commenting those variables in the config file is what causes the error[22:19:37] <grigora> though you are right it doesn't make sense[22:19:47] <grigora> Nov 2 14:12:10 tsit postfix/smtpd[19029]: warning: connect to pgsql server localhost: SSL SYSCALL error: No such file or directory?[22:20:27] <rob0> so WHY use ssl to localhost? Who's sniffing that traffic?[22:21:56] <grigora> rob0: it shouldn't ... maybe the server is configured to use SSL on all inet connections, including the localhost[22:24:39] *** growltiger has joined #postfix[22:26:31] *** Stijn has joined #postfix[22:27:24] <Stijn> Hello, I've had some trouble with my packet manager which has left me with an unusable amavisd, I would like to route my mail directly instead of through the amavisd, so I have put back the normail (my distributions) master.cf[22:27:44] <Stijn> I however still get the message 'mail transport unavailable', status = deferred[22:27:54] <Stijn> And I've got a huge postqueue :)[22:27:59] <Stijn> Any hints where to look ?[22:28:24] <Stijn> Can't really find anything in mail.log that gives anything usefull (except all the deferred messages)[22:30:33] <grigora> rob0: ssl was enabled on the database server and has now been disabled and postfix can now query fine even without the tls enabled. Thanks[22:31:02] <Stijn> ( I use postfix / maildrop )[22:37:01] <Stijn> Anyone[22:37:02] <Stijn> ?[22:38:01] <grigora> Stijn: make sure your content_filter variable in main.cf is not set to use amavis[22:39:51] <Stijn> grigora: No such thing :/[22:40:00] <Stijn> This is my master.cf: http://pastebin.archlinux.org/17144[22:40:38] *** etaylor has quit IRC[22:41:59] *** Ryushin has joined #postfix[22:42:18] <Stijn> What would be my mailtransfeR ?[22:42:21] <Stijn> maildrop ?[22:42:43] <Ryushin> Is there a way to transfer all the mail in the incoming queue to the active queue?[22:43:01] <grigora> Stijn: i was talking about main.cf not master.cf[22:43:17] <Xzisted> how would one force postfix to process all the mail in queue[22:43:24] <Ryushin> I don't see a command that will do that. Postsuper can't do that. At least not what I've been reading in the man page.[22:43:53] *** ke4qqq has quit IRC[22:44:06] <Ryushin> Xzisted: postfix flush should processes everything in the deferred queue.[22:44:12] <grigora> Xzisted: postqueue -f?[22:44:24] <Stijn> grigora: I know, the option was not in there[22:50:16] *** echelog has joined #postfix[22:50:37] <Xzisted> ok....so here is my problem. i set up all mail destined for example.com (which this server needs to send to a smart host) to go to corpmail.example.com. however, corpmail.example.com does not have an mx record....its just a processing host. so postfix cant find it. how do i get it to just blanket forward to a host without doing an mx lookup on it[22:50:45] <Stijn> and : http://opendir.stijnspijker.nl/postconf_-n.txt[22:50:50] <Stijn> rob0: There's the -n[22:51:47] <grigora> Xzisted: just use the host's IP address[22:51:52] <Ryushin> My company sends out a lot of e-mail at the beginning of each month. We are still trying to catch up from yesterday. Can someone look at my main.cf and see what I might change to speed things up. I'll put it up on pastebin[22:51:52] <Stijn> And I'm using this setup: http://gentoo-wiki.com/HOWTO_Email:_A_Complete_Virtual_System_-_Installing_Amavisd_and_SpamAssassin[22:51:53] <Stijn> without the content-filter stuff[22:53:10] *** x-ip has quit IRC[22:53:11] <Stijn> But the point is, where could the problem with the transport be, where is this configured ?[22:53:18] <Stijn> or what even is my transport ? :)[22:53:34] <Stijn> Just answering these questions would probably enable me solving the problem myself[22:54:04] <Ryushin> Xzisted: You'll need to set up a transport. Put this in main.cf: transport_maps = hash:/etc/postfix/transport Add this to transport: destinedomain.com smtp:[nameofserver.destineddomain.com][22:54:23] <Xzisted> grigora: so in my transport i would have the following: example.com smtp:corpmail.example.com[22:54:30] <Ryushin> Xzisted: run postmap /etc/postfix/transport to make a hash of it.[22:54:39] <Xzisted> Ryushin: thats exactly how i have it set up....but its doing an mx record lookup against it[22:54:44] <Ryushin> You have to put it in brackets for it not to do a mx lookup.[22:54:59] <Xzisted> Ryushin: ahh...dont see THAT in the man page[22:55:01] <Xzisted> Ryushin: thx[22:55:13] <Ryushin> Np.[22:55:29] <grigora> Stijn: so this config uses content_filter option when specifying the various transports in master.cf[22:55:45] *** hemry has quit IRC[22:56:08] *** hemry has joined #postfix[22:56:29] <Stijn> grigora: Then why is it not in my main.cf ?[22:56:48] <Stijn> And why did it work ?[22:57:01] <sysmonk> grigora just said that it's in master.cf...[22:57:15] <Stijn> indeed[22:57:26] <Stijn> My master.cf doesn't specify content_filter anymore though..[22:57:34] <Stijn> because it's reset to distribution default[22:57:39] <b4> ah sheesh, I guess I was really searching with my arse[22:57:45] <b4> found the option.[22:58:23] <grigora> Stijn: take a look at http://www.freespamfilter.org/FC4.html[22:59:06] <Stijn> I appreciate your help, but how could this link help me ?[22:59:11] <rob0> Stijn: Why are you asking *us* what you had in your config? So far, nothing you posted has shown anything which would direct mail to a content filter. Next try some logs.[22:59:49] <Stijn> Exactly![23:00:07] <Stijn> That's why I don't understand why postfix is telling me the transport is unavailable[23:00:12] <Stijn> and the logs don't say anything[23:00:14] <rob0> Exactly, as I should read your mind?[23:00:21] <Stijn> I'll search for a debug option[23:00:40] <Stijn> Since I don't refer to any transport anymore, it's basically a standard install[23:00:59] <Ryushin> Here is my main.cf: http://www.pastebin.ca/759156 Can someone smarter than me take a look at it and let me know what I can do to improve efficiency.[23:01:05] <rob0> how was it going to amavisd BEFORE? What did you change?[23:01:06] <Xzisted> Ryushin: thanks...now im getting timeouts instead of bad lookups (which is what i want...havent fixed the firewall yet)[23:01:19] <Ryushin> Xzisted: Good to hear.[23:02:33] <Stijn> rob0: As far as I know, the master.cf directed everything to the amavisd smtp daemon, it was then scanned, and resubmitted to postfix[23:02:38] *** ack_syn has joined #postfix[23:02:45] <grigora> Stijn: every entry in master.cf is a transport, so just because you are not using amavis/content_filter, doesn't mean you are not using transports[23:02:48] *** Mazon has quit IRC[23:02:55] <Stijn> rob0: I used this guide : http://gentoo-wiki.com/HOWTO_Email:_A_Complete_Virtual_System_-_Installing_Amavisd_and_SpamAssassin[23:03:18] <Stijn> But now I have placed back the standard master.cf, so it should not use amavisd anymore[23:03:21] <grigora> Stijn: go to your master.cf and enable -v next to smtpd in the rightmost column[23:03:39] <Stijn> I was just going to do that :)[23:03:41] <grigora> Stijn: just a silly question, have you restarted your server?[23:04:32] <Stijn> yes[23:04:51] <Stijn> Because my packetmanager failed miserably after downgrading some packages[23:04:58] <Stijn> Helas, I don't have physical access for two months[23:05:09] <Stijn> So I can't reinstall it[23:05:18] <grigora> Stijn: also check your /etc/postfix/transport[23:05:26] <Stijn> grigora: It only contains the man page[23:05:34] <sepski> Stijn, what distro ?[23:05:37] <Stijn> archlinux[23:05:56] <sepski> deb or rpm based ?[23:06:00] <Stijn> none[23:06:03] <Stijn> own packetmanager[23:06:05] <Stijn> pacman :)[23:06:08] <sepski> right then i have no clue[23:06:26] <sepski> if it was debian based you'd find the packetmanager backups in /var/backups :P[23:06:26] <Stijn> pacman is very nice, except for downgrading, as I have discovered, it does not support that[23:06:39] <grigora> Stijn: i haven't looked at this in a while, but maybe when you specify the variable, the server expects a valid transport_map ...[23:06:44] <Stijn> because archlinux is a rolling release[23:06:58] *** salryman has joined #postfix[23:07:48] <Stijn> grigora: I don't seem to be getting any more debugging in logs after setting smtp to -v, and I did a postfix restart && postqueue -f[23:09:06] <grigora> just cut and paste the single line from the logs containing the error message[23:09:27] <Stijn> Nov 2 23:08:10 rewurst postfix/error[3933]: F041036322: to=<stijn at stijnspijker dot nl>, relay=none, delay=369181, delays=369180/0.41/0/0, dsn=4.3.0, status=deferred (mail transport unavailable)[23:10:39] <grigora> Stijn: is this incoming or outgoing?[23:10:57] <Stijn> these are incoming[23:11:11] <Stijn> I did have a view of these about a day ago:[23:11:11] <Stijn> Nov 2 21:59:57 rewurst postfix/qmgr[8595]: F1A073631D: to=<stijn at stijnspijker dot nl>, relay=none, delay=382853, delays=382853/0/0/0, dsn=4.4.1, status=deferred (delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]: Connection refused)[23:11:27] <Stijn> Well, and one today ^-^[23:11:52] <grigora> Stijn: make sure your local delivery is working, I think you mentioned you were using maildrop[23:11:57] <Stijn> yes[23:12:19] <Stijn> How could I test it ?[23:12:30] <Stijn> executing it works, so :p[23:12:33] <grigora> i don't know maildrop[23:12:42] <grigora> but your maildrop line in master.cf is commented out[23:14:21] <Stijn> Uncommented it, but that didn't work :/[23:14:59] <grigora> it may involve a little more than uncommenting the line, just look in the maildrop docs and do it right[23:15:22] <Stijn> k[23:15:43] <grigora> my guess is that your local delivery is broken, postfix doesn't know how to delivery the mail to you once it receives it[23:15:50] *** n3kl has quit IRC[23:15:56] <grigora> Stijn: good luck[23:15:58] *** grigora has left #postfix[23:19:51] *** solar_ant has quit IRC[23:22:21] *** havvg_ has joined #postfix[23:23:49] *** felix_da_catz has joined #postfix[23:27:43] *** pitakill has quit IRC[23:36:54] *** havvg has quit IRC[23:37:47] *** davedoom2 has joined #postfix[23:37:50] <davedoom2> hi![23:38:08] *** sepski has quit IRC[23:38:37] <davedoom2> does anyone know of a program to convert qmail aliases to postfix format?[23:39:23] <rob0> I can't remember the qmail format.[23:40:06] <davedoom2> it is pretty involved[23:40:19] *** havvg_ has quit IRC[23:40:26] <davedoom2> i will write a shell script to convert if no one has done it before[23:46:00] *** collink_ has left #postfix[23:52:36] *** ack_syn has quit IRC[23:53:12] <Stijn> :/[23:53:19] <Stijn> Can't get it to work[23:53:23] <Stijn> too little debug info[23:55:41] *** hemry has quit IRC[23:56:43] <Stijn> Can anyone do something with this debug message:[23:56:44] <Stijn> Nov 2 23:55:40 rewurst postfix/qmgr[4759]: qmgr_queue_unthrottle: queue 4.3.0 mail transport unavailable[23:58:34] <Stijn> AHA ! :)[23:59:10] <Stijn> Can anyone help me with this one:[23:59:11] <Stijn> Nov 2 23:58:39 rewurst postfix/qmgr[4853]: warning: connect to transport scan: Connection refused[23:59:19] <Stijn> How can I disable the scan transport ?[23:59:22] <Stijn> it's not in master.cf[23:59:38] <Stijn> should I add it to master.cf and redirect it to something ?