Archive-Date: Thu, 5 Aug 1999 06:26:05 -0700 Sender: system@mailgate.spotimage.fr Date: Thu, 05 Aug 1999 13:27:57 GMT From: System Manager Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Message-ID: <009DC2CE.CB1E8700.17@spotimage.fr> Subject: local error is lost $ mcp show path Domain-to-path mappings: Domain="mailhub.spotimage.int", Path=SMTP Domain="mailgate.spotimage.fr", Path=Local Domain="mailgate", Path=Local Domain="*", Path=SMTP, Route="[194.199.150.1]" $ mcp show rewrite Address-rewriting rules: Rewrite "<@{route}:{local}@{domain}>" => "<@{route}:{local}@{domain}>" Rewrite "<{first}.{last}@spotimage.{domain}>" => "<@mailhub.spotimage.int:{first}.{last}@spotimage.fr>" Rewrite "<{username}@spotimage.{domain}>" => "<{username}@mailgate.spotimage.fr>" with the above paths/rewrite, if I send a mail to mx%foo, I receive no error message. I should as user foo does not exists. here are the corresponding log files. $ type MX_ROOT:[ROUTER]MX_ROUTER_LOG.LOG;7 5-AUG-1999 13:24:55.89 %PROCESS, Processing entry number 87 5-AUG-1999 13:24:55.94 %PROCESS, Status from READ_INFO was 00000001 5-AUG-1999 13:24:55.95 %PROCESS, Message originated in VMS Mail. 5-AUG-1999 13:24:55.96 %PROCESS, will run domain expander on envelope addresses. 5-AUG-1999 13:24:55.96 %PROCESS, Processing address: 5-AUG-1999 13:24:55.96 %PROCESS, ... address now reads: 5-AUG-1999 13:24:55.96 %PROCESS, will run domain expander on message headers. 5-AUG-1999 13:24:56.28 %PROCESS, Updating the QENT source address. 5-AUG-1999 13:24:56.30 %PROCESS, Finished VMSmail-origin preprocessing. 5-AUG-1999 13:24:56.30 %PROCESS, Recipient #0: 5-AUG-1999 13:24:56.30 %REWRITE, No rewrite rules matched 5-AUG-1999 13:24:56.30 %FINDPATH, Site-spec expand on mailgate.spotimage.fr err=00000000 5-AUG-1999 13:24:56.30 %FINDPATH, domain name MAILGATE.SPOTIMAGE.FR matched path pattern MAILGATE.SPOTIMAGE.FR 5-AUG-1999 13:24:56.30 %PROCESS, Rewrote as - next hop mailgate.spotimage.fr, path 1 5-AUG-1999 13:24:56.31 %FINDALIAS, no alias found for foo 5-AUG-1999 13:24:56.31 %PROCESS, no alias found for foo 5-AUG-1999 13:24:56.31 %PROCESS, this is just a local delivery 5-AUG-1999 13:24:56.37 %PROCESS, Adding to LOCAL path: foo. 5-AUG-1999 13:24:56.49 %PROCESS, Path LOCAL gets 1 rcpts, entry number 88 $ type MX_ROOT:[LOCAL]MX_LOCAL_LOG.LOG;7 5-AUG-1999 13:24:56.64 Processing queue entry number 88 5-AUG-1999 13:24:56.70 Checking local name: FOO 5-AUG-1999 13:24:56.71 Don't know what this is (status = 000182B2). 5-AUG-1999 13:24:57.26 All done with this entry. $ exit %X000182B2 %RMS-E-RNF, record not found TIA, Pierre. ================================================================================ Archive-Date: Thu, 5 Aug 1999 06:48:19 -0700 Sender: madison@MadGoat.Com Date: Thu, 5 Aug 1999 06:48:13 -0700 From: Matt Madison Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: system@spotimage.fr Message-ID: <009DC296.F3AA9B24.22@MadGoat.Com> Subject: RE: local error is lost [...] >with the above paths/rewrite, if I send a mail to mx%foo, I receive no error message. >I should as user foo does not exists. Do you have any debug log files for the bounce message? It's possible that a message is getting generated, but somehow isn't getting back to you. -Matt -- Matthew Madison | MadGoat Software | PO Box 556, Santa Cruz, CA 95061 USA madison@madgoat.com http://www.madgoat.com ================================================================================ Archive-Date: Thu, 5 Aug 1999 06:58:27 -0700 Sender: system@mailgate.spotimage.fr Date: Thu, 05 Aug 1999 14:00:20 GMT From: System Manager Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: madison@MadGoat.com Message-ID: <009DC2D3.5199E0A0.22@spotimage.fr> Subject: RE: local error is lost >Do you have any debug log files for the bounce message? It's possible that a message >is getting generated, but somehow isn't getting back to you. oops...sorry... mail with 'local delivery error' in the subject are redirected (with DELIVER%) from the postmaster account to another account and I did my testings from the postamster account. so both messages (the one for the originator and the copy for the postmaster) were redirected to the other account. sorry again. Pierre. ================================================================================ Archive-Date: Fri, 6 Aug 1999 09:26:07 -0700 Subject: MX_SMTP_LOCK_EXCLUSIONS again From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37ab0b4b@news.kapsch.co.at> Date: 6 Aug 1999 18:20:27 +0100 To: MX-List@LISTS.WKU.EDU In MX V5.1 there is a (undocumented) logical MX_SMTP_LOCK_EXCLUSIONS From an recent example: $ DEFINE /SYSTEM /EXEC MX_SMTP_LOCK_EXCLUSIONS - "xxx.domain.com",- "172.27.11.23",- "yyy.domain.com",- "192.168.234.4" I tried this for my domains with $ DEF/SYS/EXE MX_SMTP_LOCK_EXCLUSIONS "*.kapsch.co.at", - "*.atc.co.at" did a $ MCP RESET/CLU SMTP. But no luck. Nodes get locked again. Do I need to define "nodename.kapsch.co.at" for every node instead ? Is the translation string (list) *) a (list of) domains ? *) a (list of) mailserver(s) for a domain ? *) allowed to contain wildcards ? Is is possible to revert the (undocumented) Lock Feature to the MX V4 behaviour with $ DEF/SYS/EXE MX_SMTP_LOCK_EXCLUSIONS "*" ? TIA -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Fri, 6 Aug 1999 12:28:24 -0700 Sender: madison@MadGoat.Com Date: Fri, 6 Aug 1999 12:28:18 -0700 From: Matt Madison Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: eplan@kapsch.net Message-ID: <009DC38F.A0C86419.12@MadGoat.Com> Subject: RE: MX_SMTP_LOCK_EXCLUSIONS again >In MX V5.1 there is a (undocumented) logical > > MX_SMTP_LOCK_EXCLUSIONS > >From an recent example: > > $ DEFINE /SYSTEM /EXEC MX_SMTP_LOCK_EXCLUSIONS - > "xxx.domain.com",- > "172.27.11.23",- > "yyy.domain.com",- > "192.168.234.4" > > >I tried this for my domains with > >$ DEF/SYS/EXE MX_SMTP_LOCK_EXCLUSIONS "*.kapsch.co.at", - > "*.atc.co.at" > >did a $ MCP RESET/CLU SMTP. But no luck. Nodes get locked again. > >Do I need to define "nodename.kapsch.co.at" for every node instead ? Yes. >Is the translation string (list) > >*) a (list of) domains ? Each string is a single domain name or IP address. >*) a (list of) mailserver(s) for a domain ? Yes, the domain name is the name of the host to which MX would actually connect. >*) allowed to contain wildcards ? No. >Is is possible to revert the (undocumented) Lock Feature to the >MX V4 behaviour with $ DEF/SYS/EXE MX_SMTP_LOCK_EXCLUSIONS "*" ? No, since wildcard matching is not used. -Matt -- Matthew Madison | MadGoat Software | PO Box 556, Santa Cruz, CA 95061 USA madison@madgoat.com http://www.madgoat.com ================================================================================ Archive-Date: Tue, 10 Aug 1999 06:22:51 -0700 Date: Tue, 10 Aug 1999 09:22:42 -0400 From: mlbowden@LIB1.HACC.EDU Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: mlbowden@LIB1.HACC.EDU Message-ID: <009DC69A.5C8BE344.113@LIB1.HACC.EDU> Subject: Preventing bounced when host is down Good morning: Our campus mail system is groupwise. (Don't ask me why. I had no say.) The groupwise server goes down regularly. When the groupwise server is down, my Alpha accepts the campus mail. The problem I am having is that after 12 retries, my Alpha bounces the messages. Is there a way to make mx hold these message indefinitely? Our groupwise server is not accepting mail and our groupwise administrator is on vacation. I have tried adjusting the SMTP DNS failure and the maxium retries. But they don't seem to help. Thanks for the help Michael L. Bowdn Harrisburg Area Community College mlbowden@hacc.edu ================================================================================ Archive-Date: Tue, 10 Aug 1999 06:28:00 -0700 Date: Tue, 10 Aug 1999 8:27:46 -0500 From: Hunter Goatley Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: mlbowden@LIB1.HACC.EDU Message-ID: <990810082746.202000c1@goat.process.com> Subject: RE: Preventing bounced when host is down mlbowden@LIB1.HACC.EDU writes: > >I have tried adjusting the SMTP DNS >failure and the maxium retries. But they don't seem to help. > If they're bouncing after 12 retries, the SMTP DNS failure retries is the key. Did you RESET the SMTP agents after changing the value? Hunter ------ Hunter Goatley, Process Software, http://www.process.com/ http://www2.wku.edu/hunter/ ================================================================================ Archive-Date: Tue, 10 Aug 1999 12:51:48 -0700 Subject: MX051 and DSN From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37b08044@news.kapsch.co.at> Date: 10 Aug 1999 21:40:52 +0100 To: MX-List@LISTS.WKU.EDU What is the default of the logical MX_VMSMAIL_DSN_CONTROL ? We didn't define it so far and we got some DSN Mails with body so far. So I would assume, it is "NOTIFY=SUCCESS RETURN=FULL". Or what ? Why isn't "NOTIFY=NEVER" the default ? Or Is the logical only for a DSN setting override and if the logical is not there, then the requesting system (who is the requesting system, my exchange servers sending the mail to MX, or is it the outside mail server finally getting the mail from MX ?) decides if MX has to [not] send a DSN ? Seems I need a little bit DSN training... Many TIA -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Tue, 10 Aug 1999 14:07:42 -0700 Date: Tue, 10 Aug 1999 14:07:33 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Subject: Re: MX051 and DSN In-Reply-To: <37b08044@news.kapsch.co.at> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On 10 Aug 1999 21:40 +0100, Peter LANGSTOEGER wrote: > What is the default of the logical MX_VMSMAIL_DSN_CONTROL ? > We didn't define it so far and we got some DSN Mails with body so far. > > So I would assume, it is "NOTIFY=SUCCESS RETURN=FULL". > Or what ? > > Why isn't "NOTIFY=NEVER" the default ? > > Or > > Is the logical only for a DSN setting override and if the logical is not > there, then the requesting system (who is the requesting system, my > exchange servers sending the mail to MX, or is it the outside mail server > finally getting the mail from MX ?) decides if MX has to [not] send a DSN ? > > > Seems I need a little bit DSN training... All of the MX_VMSMAIL_* logicals only apply to messages sent using VMSmail using MX: $ MAIL MAIL> SEND To: MX%"blah@blather" -Dan Wing ================================================================================ Archive-Date: Tue, 10 Aug 1999 14:50:47 -0700 Subject: Re: MX051 and DSN From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37b09b46@news.kapsch.co.at> Date: 10 Aug 1999 23:36:06 +0100 To: MX-List@LISTS.WKU.EDU In article , Dan Wing writes: >All of the MX_VMSMAIL_* logicals only apply to messages sent using >VMSmail using MX: > > $ MAIL > MAIL> SEND > To: MX%"blah@blather" Thanks. I just noticed it. I did a $ DEF/SYS/EXE MX_VMSMAIL_DSN_CONTROL "NOTIFY=NEVER" and now I get MAIL> s To: blabla %MX-I-BADDSNCTL, DSN control settings are invalid and have been ignored There seems to be a bug somewhere... And so the other question: Is there a way to disable DSN on MX permanently or is this a job of the other mail servers (especially our exchange servers in the local branches which seems to get some big DSN mails with full mail content over a slow WAN link) ? TIA -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Tue, 10 Aug 1999 15:45:07 -0700 Date: Tue, 10 Aug 1999 15:44:57 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Subject: Re: MX051 and DSN In-Reply-To: <37b09b46@news.kapsch.co.at> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On 10 Aug 1999 23:36 +0100, Peter LANGSTOEGER wrote: > In article , Dan Wing writes: > >All of the MX_VMSMAIL_* logicals only apply to messages sent using > >VMSmail using MX: > > > > $ MAIL > > MAIL> SEND > > To: MX%"blah@blather" > > Thanks. > I just noticed it. I did a > > $ DEF/SYS/EXE MX_VMSMAIL_DSN_CONTROL "NOTIFY=NEVER" > > and now I get > > MAIL> s > To: blabla > %MX-I-BADDSNCTL, DSN control settings are invalid and have been ignored > > There seems to be a bug somewhere... Yeah, I can duplicate that error, but it only seems to happen with the setting "NOTIFY=NEVER", but the error doesn't happen with other valid values for that logical such as "NOTIFY=SUCCESS", "NOTIFY=SUCCESS,FAILURE", or "NOTIFY=SUCCESS,FAILURE,DELAY". Note you NEVER want "NOTIFY=NEVER" anyway. Such a setting would prevent bounces such as "no such user" and "invalid domain" from being sent. This is usually never desired and is akin to sending all of your correspondence in the US mail without any return address. > And so the other question: > > Is there a way to disable DSN on MX permanently or is this a job of > the other mail servers (especially our exchange servers in the local > branches which seems to get some big DSN mails with full mail content > over a slow WAN link) ? No, there isn't. If MX is asked to do DSN processing, MX will do it. Some various Microsoft mailers misbehave with RFC1891-compliant mailers which is I'm sure what you're seeing. It is the job of whatever mailer is submitting the mail to its mail submission server to request the type of DSN it desires. Non-DSN-aware mailers have the functionally equivalent setting to "NOTIFY=FAILURE" and the headers-only/full-message decision is made by the mailer generating the NDN (non delivery notification). -d ================================================================================ Archive-Date: Wed, 11 Aug 1999 07:07:36 -0700 Sender: madison@MadGoat.Com Date: Wed, 11 Aug 1999 07:07:30 -0700 From: Matt Madison Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: eplan@kapsch.net Message-ID: <009DC750.A4002481.15@MadGoat.Com> Subject: Re: MX051 and DSN >In article , Dan Wing writes: >>All of the MX_VMSMAIL_* logicals only apply to messages sent using >>VMSmail using MX: >> >> $ MAIL >> MAIL> SEND >> To: MX%"blah@blather" > >Thanks. >I just noticed it. I did a > >$ DEF/SYS/EXE MX_VMSMAIL_DSN_CONTROL "NOTIFY=NEVER" > >and now I get > >MAIL> s >To: blabla >%MX-I-BADDSNCTL, DSN control settings are invalid and have been ignored > >There seems to be a bug somewhere... Yup, there is. I can issue a patch if anyone's interested, otherwise it will be fixed in the next release. However, as Dan pointed out, NOTIFY=NEVER is generally not a good setting to use (and is the only setting affected by the parsing bug). >Is there a way to disable DSN on MX permanently or is this a job of the other >mail servers (especially our exchange servers in the local branches which >seems to get some big DSN mails with full mail content over a slow WAN link) ? It is the responsibility of the sending system to request DSNs, which MX honors. I did put in a change in ECO 4 that allows you to disable the "delayed message" notifications, which someone else asked for who had uncooperative Exchange servers, but there's no way to turn of DSN support completely. I don't know whether Exchange's DSN usage can be configured. -Matt -- Matthew Madison | MadGoat Software | PO Box 556, Santa Cruz, CA 95061 USA madison@madgoat.com http://www.madgoat.com ================================================================================ Archive-Date: Thu, 12 Aug 1999 02:46:50 -0700 Subject: Re: MX051 and DSN From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37b29700@news.kapsch.co.at> Date: 12 Aug 1999 11:42:24 +0100 To: MX-List@LISTS.WKU.EDU In article <009DC750.A4002481.15@MadGoat.Com>, Matt Madison writes: >It is the responsibility of the sending system to request DSNs, which MX >honors. I did put in a change in ECO 4 that allows you to disable the >"delayed message" notifications, which someone else asked for who had >uncooperative Exchange servers, but there's no way to turn of DSN support >completely. I did already install ECOs (up to 10) and I know of and have defined the logical MX_DSN_DISABLE_INITIAL_DELAY_NOTIFICATIONS. I also defined MX_DSN_NOTIFICATION_INTERVAL with "0 12:00:00". But in my case, Delay Notifications have never been a problem. DSN support have been added in MX5. The only chance to disable DSN would then be downgrading to MX4 and losing many features (Anti-SPAM) and bugfixes ? TIA -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Thu, 12 Aug 1999 08:50:14 -0700 Date: Thu, 12 Aug 1999 08:50:06 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: MX-List@LISTS.WKU.EDU Subject: Re: MX051 and DSN In-Reply-To: <37b29700@news.kapsch.co.at> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On 12 Aug 1999 11:42 +0100, Peter LANGSTOEGER wrote: > In article <009DC750.A4002481.15@MadGoat.Com>, Matt Madison writes: > >It is the responsibility of the sending system to request DSNs, which MX > >honors. I did put in a change in ECO 4 that allows you to disable the > >"delayed message" notifications, which someone else asked for who had > >uncooperative Exchange servers, but there's no way to turn of DSN support > >completely. > > I did already install ECOs (up to 10) and I know of and have defined the > logical MX_DSN_DISABLE_INITIAL_DELAY_NOTIFICATIONS. > I also defined MX_DSN_NOTIFICATION_INTERVAL with "0 12:00:00". > > But in my case, Delay Notifications have never been a problem. > > DSN support have been added in MX5. The only chance to disable DSN would then > be downgrading to MX4 and losing many features (Anti-SPAM) and bugfixes ? It sounds like you're wanting DSN modified such so that when a certain Microsoft mailer asks for a DSN you will appease the mailer and make it think you're going to Do The Right Thing but in reality you don't want the DSN request forwarded to the next mailer because it causes too much traffic? This breaks down the premise of DSNs which are a hop-by-hop guarantee to provide delivery status indications to the sender when the sender wanted such indications. The solution in other mailers that I use that support DSN (pine, Netscape) is to simply configure the mailer to not request DSNs. Is the Microsoft product unable to turn off its request for DSNs? I believe the reason this happens is that some Microsoft mailer doesn't honor requests for DSNs. I know the mailer at microsoft.com itself behaved this way previously and I don't know if MS was planning on implementing the RFC as it is written or continuing on their path of doing their own thing and forcing folks to buy non-compliant software across the board to avoid problems such as what you are encountering. -Dan Wing ================================================================================ Archive-Date: Thu, 12 Aug 1999 08:50:21 -0700 Date: Thu, 12 Aug 1999 08:50:06 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: MX-List@LISTS.WKU.EDU Subject: Re: MX051 and DSN In-Reply-To: <37b29700@news.kapsch.co.at> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On 12 Aug 1999 11:42 +0100, Peter LANGSTOEGER wrote: > In article <009DC750.A4002481.15@MadGoat.Com>, Matt Madison writes: > >It is the responsibility of the sending system to request DSNs, which MX > >honors. I did put in a change in ECO 4 that allows you to disable the > >"delayed message" notifications, which someone else asked for who had > >uncooperative Exchange servers, but there's no way to turn of DSN support > >completely. > > I did already install ECOs (up to 10) and I know of and have defined the > logical MX_DSN_DISABLE_INITIAL_DELAY_NOTIFICATIONS. > I also defined MX_DSN_NOTIFICATION_INTERVAL with "0 12:00:00". > > But in my case, Delay Notifications have never been a problem. > > DSN support have been added in MX5. The only chance to disable DSN would then > be downgrading to MX4 and losing many features (Anti-SPAM) and bugfixes ? It sounds like you're wanting DSN modified such so that when a certain Microsoft mailer asks for a DSN you will appease the mailer and make it think you're going to Do The Right Thing but in reality you don't want the DSN request forwarded to the next mailer because it causes too much traffic? This breaks down the premise of DSNs which are a hop-by-hop guarantee to provide delivery status indications to the sender when the sender wanted such indications. The solution in other mailers that I use that support DSN (pine, Netscape) is to simply configure the mailer to not request DSNs. Is the Microsoft product unable to turn off its request for DSNs? I believe the reason this happens is that some Microsoft mailer doesn't honor requests for DSNs. I know the mailer at microsoft.com itself behaved this way previously and I don't know if MS was planning on implementing the RFC as it is written or continuing on their path of doing their own thing and forcing folks to buy non-compliant software across the board to avoid problems such as what you are encountering. -Dan Wing ================================================================================ Archive-Date: Fri, 13 Aug 1999 02:37:50 -0700 Subject: Why %MX-W-HOSTLOCK ? From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37b3e753@news.kapsch.co.at> Date: 13 Aug 1999 11:37:23 +0100 To: MX-List@LISTS.WKU.EDU I've a mail pending in the MX queue to a mail address which has neither an A RR nor an MX RR. I would expect to see a message like "no mailserver availablable" or similar. But the message is %MX-W-HOSTLOCK, target host ...". I wonder why. What node is MX trying to reach (which seems unreachable then) ? TIA -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Fri, 13 Aug 1999 05:08:00 -0700 Date: Fri, 13 Aug 1999 7:07:51 -0500 From: Hunter Goatley Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Message-ID: <990813070751.202000c1@process.com> Subject: RE: Why %MX-W-HOSTLOCK ? eplan@kapsch.net (Peter LANGSTOEGER) writes: > >I've a mail pending in the MX queue to a mail address which has neither >an A RR nor an MX RR. I would expect to see a message like "no mailserver >availablable" or similar. But the message is %MX-W-HOSTLOCK, target host ...". > That's because it has tried multiple times and failed, so the files in MX_SMTP_LOCK_DIR: are being used now. >I wonder why. What node is MX trying to reach (which seems unreachable then) ? > Turn on SMTP debugging and READY an entry.... Hunter ------ Hunter Goatley, Process Software, http://www.process.com/ http://www2.wku.edu/hunter/ ================================================================================ Archive-Date: Fri, 13 Aug 1999 05:49:14 -0700 Sender: jtb@ATM.COM.PL Message-ID: <99Aug13.145422met_dst.14339@gateway.hq.atm.com.pl> Date: Fri, 13 Aug 1999 14:42:58 +0200 From: Jacek Tobiasz Reply-To: MX-List@MadGoat.com MIME-Version: 1.0 To: MX-List@MadGoat.com Subject: Re: Why %MX-W-HOSTLOCK ? References: <990813070751.202000c1@process.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hunter Goatley wrote: > > eplan@kapsch.net (Peter LANGSTOEGER) writes: > > > >I've a mail pending in the MX queue to a mail address which has neither > >an A RR nor an MX RR. I would expect to see a message like "no mailserver > >availablable" or similar. But the message is %MX-W-HOSTLOCK, target host ...". > > > That's because it has tried multiple times and failed, so the files in > MX_SMTP_LOCK_DIR: are being used now. Yes, I wanted to write the same, but ... yes, it got DNS response - no such host/domain. Is MX able to recongize when DNS server is down (no answer) and when it gets 'no such host' response ? Or is it more complicated ? Regards Jacek ================================================================================ Archive-Date: Fri, 13 Aug 1999 05:54:02 -0700 Date: Fri, 13 Aug 1999 7:53:52 -0500 From: Hunter Goatley Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Message-ID: <990813075352.202000c1@process.com> Subject: Re: Why %MX-W-HOSTLOCK ? Jacek Tobiasz writes: > >Yes, I wanted to write the same, but ... yes, it got DNS response - no such >host/domain. Is MX able to recongize when DNS server is down (no answer) >and when it gets 'no such host' response ? > Yes, it makes the distinction, but it doesn't affect the number of retries, if I remember correctly. Hunter ------ Hunter Goatley, Process Software, http://www.process.com/ http://www2.wku.edu/hunter/ ================================================================================ Archive-Date: Fri, 13 Aug 1999 06:56:58 -0700 Subject: RE: Why %MX-W-HOSTLOCK ? From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37b4236f@news.kapsch.co.at> Date: 13 Aug 1999 15:53:51 +0100 To: MX-List@LISTS.WKU.EDU In article <990813070751.202000c1@process.com>, Hunter Goatley writes: >eplan@kapsch.net (Peter LANGSTOEGER) writes: >>I've a mail pending in the MX queue to a mail address which has neither >>an A RR nor an MX RR. I would expect to see a message like "no mailserver >>availablable" or similar. But the message is %MX-W-HOSTLOCK, target host ...". >> >That's because it has tried multiple times and failed, so the files in >MX_SMTP_LOCK_DIR: are being used now. This is news to me. In the Lock file there is the line 13-AUG-1999 15:11:14.77 0C2783B8: %MX-W-NOCONTACT, could not establish contact with any mail servers for this destination This is the message I previously always saw with MCP QUE SHOW/FULL and this is the line I was missing. IIRC There haven't been a lock file for this situation so far. When (and why) did this behaviour change ? I recently installed ECO 6-10 (I used to have ECO 1-5). Is this the cause ? >>I wonder why. What node is MX trying to reach (which seems unreachable then) ? >> >Turn on SMTP debugging and READY an entry.... No entry in debug file (as expected). I think, when the message is sent back (because undeliverable) to the sender then the message included the %MX-W-NOCONTACT message and contains now the %MX-W-HOSTLOCK message. The user now no longer knows why the message did not reach the recipient. Correct ? TIA -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Fri, 13 Aug 1999 07:04:54 -0700 Date: Fri, 13 Aug 1999 9:04:44 -0500 From: Hunter Goatley Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Message-ID: <990813090444.202000c1@process.com> Subject: RE: Why %MX-W-HOSTLOCK ? eplan@kapsch.net (Peter LANGSTOEGER) writes: > >>That's because it has tried multiple times and failed, so the files in >>MX_SMTP_LOCK_DIR: are being used now. > >This is news to me. > A more thorough explanation is probably in order. When a failure for a host occurs, a file is created in MX_SMTP_LOCK_DIR: to prevent the other SMTP agents from attempting to contact that host within a certain period of time. Before this was done, you could have 20 SMTP agents all trying to deliver mail to the same unreachable host one right after the other. >In the Lock file there is the line > >13-AUG-1999 15:11:14.77 0C2783B8: %MX-W-NOCONTACT, could not establish contact >with any mail servers for this destination > >This is the message I previously always saw with MCP QUE SHOW/FULL and >this is the line I was missing. >IIRC There haven't been a lock file for this situation so far. > >When (and why) did this behaviour change ? >I recently installed ECO 6-10 (I used to have ECO 1-5). Is this the cause ? > No, it's always been in V5.x (unless there was a bug I don't remember). Unless you read the docs, you wouldn't necessarily know the files were there.... >>>I wonder why. What node is MX trying to reach (which seems unreachable then) ? >>> >>Turn on SMTP debugging and READY an entry.... > >No entry in debug file (as expected). > Right---because the file is present. Delete the MX_SMTP_LOCK_DIR: file(s) for that host, then try it again. You should see a debug entry this time. >I think, when the message is sent back (because undeliverable) to the sender >then the message included the %MX-W-NOCONTACT message and contains now the >%MX-W-HOSTLOCK message. The user now no longer knows why the message did not >reach the recipient. Correct ? > Yes, that is true. Hunter ------ Hunter Goatley, Process Software, http://www.process.com/ http://www2.wku.edu/hunter/ ================================================================================ Archive-Date: Fri, 13 Aug 1999 07:41:43 -0700 Subject: RE: Why %MX-W-HOSTLOCK ? From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37b42d64@news.kapsch.co.at> Date: 13 Aug 1999 16:36:20 +0100 To: MX-List@LISTS.WKU.EDU In article <990813090444.202000c1@process.com>, Hunter Goatley writes: >eplan@kapsch.net (Peter LANGSTOEGER) writes: >>>That's because it has tried multiple times and failed, so the files in >>>MX_SMTP_LOCK_DIR: are being used now. >> >>This is news to me. >> >A more thorough explanation is probably in order. When a failure for >a host occurs, a file is created in MX_SMTP_LOCK_DIR: to prevent the >other SMTP agents from attempting to contact that host within a >certain period of time. Before this was done, you could have 20 SMTP >agents all trying to deliver mail to the same unreachable host one >right after the other. Sorry. I do now know lock files. I know what they are for. I meant: It is news for me, that HOST_LOCK (and ADDRESS_LOCK ?) files are now created for a mail to a domain which has no MX and no A resource record. And I don't see, any advantage on this new behaviour. Only disadvantages. >>In the Lock file there is the line >> >>13-AUG-1999 15:11:14.77 0C2783B8: %MX-W-NOCONTACT, could not establish contact >>with any mail servers for this destination >> >>This is the message I previously always saw with MCP QUE SHOW/FULL and >>this is the line I was missing. >>IIRC There haven't been a lock file for this situation so far. >> >>When (and why) did this behaviour change ? >>I recently installed ECO 6-10 (I used to have ECO 1-5). Is this the cause ? >> >No, it's always been in V5.x (unless there was a bug I don't remember). A bug ? I think different. Now it is a bug and previously there wasn't one. >Unless you read the docs, you wouldn't necessarily know the files were >there.... I must admit, I haven't checked for lock files for this situation before. But I definitely remember seeing the MX-W-NOCONTACT message in QUEUE SHOW and now I see MX-W-HOSTLOCK messages instead. >>No entry in debug file (as expected). >> >Right---because the file is present. Delete the MX_SMTP_LOCK_DIR: >file(s) for that host, then try it again. You should see a debug >entry this time. Yes, there is one. Recipient: , route=domain.name SMTP_SEND: looking up host name domain.name SMTP_SEND: NETLIB_DNS_MX_LOOKUP status is 00000001 But that is all. No connection attempt. Seems ok. (Where to connect to ?) btw: Why is there already a route known when the Lookup occurs one line later ? >>I think, when the message is sent back (because undeliverable) to the sender >>then the message included the %MX-W-NOCONTACT message and contains now the >>%MX-W-HOSTLOCK message. The user now no longer knows why the message did not >>reach the recipient. Correct ? >> >Yes, that is true. So, some time before, the user did get the message %MX-W-NOCONTACT instead of %MX-W-HOSTLOCK. And this means, we have now a bug. We had previously V5.1 and now have V5.1. Please clarify. I thought a HOST_LOCK and ADDRESS_LOCK is there for specific mailservers currently not reachable. But in this situation, there is no mailserver for this domain at all. -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Fri, 13 Aug 1999 07:46:33 -0700 Date: Fri, 13 Aug 1999 9:46:24 -0500 From: Hunter Goatley Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Message-ID: <990813094624.202000c1@process.com> Subject: RE: Why %MX-W-HOSTLOCK ? eplan@kapsch.net (Peter LANGSTOEGER) writes: > >So, some time before, the user did get the message %MX-W-NOCONTACT instead of >%MX-W-HOSTLOCK. And this means, we have now a bug. We had previously V5.1 >and now have V5.1. > >Please clarify. > You'll have to wait for Matt to answer this one, as all of that is code he added that I'm not that familiar with. >I thought a HOST_LOCK and ADDRESS_LOCK is there for specific mailservers >currently not reachable. But in this situation, there is no mailserver for >this domain at all. > I suspect MX is treating all of these cases as the same. Hunter ------ Hunter Goatley, Process Software, http://www.process.com/ http://www2.wku.edu/hunter/ ================================================================================ Archive-Date: Fri, 13 Aug 1999 07:58:59 -0700 Sender: madison@MadGoat.Com Date: Fri, 13 Aug 1999 07:58:53 -0700 From: Matt Madison Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Message-ID: <009DC8EA.266DFB6A.45@MadGoat.Com> Subject: RE: Why %MX-W-HOSTLOCK ? >>So, some time before, the user did get the message %MX-W-NOCONTACT instead of >>%MX-W-HOSTLOCK. And this means, we have now a bug. We had previously V5.1 >>and now have V5.1. >> >>Please clarify. >> >You'll have to wait for Matt to answer this one, as all of that is >code he added that I'm not that familiar with. I don't think anything changed in that area, but I'll take a look at the code over the weekend and see if I can find something. -Matt -- Matthew Madison | MadGoat Software | PO Box 556, Santa Cruz, CA 95061 USA madison@madgoat.com http://www.madgoat.com ================================================================================ Archive-Date: Mon, 16 Aug 1999 22:36:11 -0700 Sender: najman@HADASSAH.org.IL Date: Tue, 17 Aug 1999 08:34:27 +0200 From: "Najman Kahana. Ext:77313" Reply-To: MX-List@MadGoat.com To: MX-LIST@MADGOAT.COM CC: najman@HADASSAH.org.IL Message-ID: <009DCC13.C7EB5CD7.14@HADASSAH.org.IL> Subject: trace sending computer Hello. I have a rather urgent, and delicate problem. I have a user who claims that someone else is sending very nasty mail under his name. To do this, is child's play. Question: is there any way of seeing the name of the source computer ? (The person whom I suspect would know how to change the username in the PC mailer, but not how to change his PC's tcp name) thank you Najman Kahana System Manager Hadassah Hospital, Israel Najman@hadassah.org.il +--------+------------------+------------------------------+---------------+ ! NAJMAN KAHANA ! Hadassah University Hospital ! thanks, ! ! Najman@hadassah.org.il ! Jerusalem, Israel ! we have our ! ! ! (visit our capital soon) ! own viruses ! +--------+------------------+------------------------------+---------------+ ================================================================================ Archive-Date: Tue, 17 Aug 1999 02:43:37 -0700 Sender: jtb@ATM.COM.PL Message-ID: <99Aug17.115158met_dst.14340@gateway.hq.atm.com.pl> Date: Tue, 17 Aug 1999 11:39:27 +0200 From: Jacek Tobiasz Reply-To: MX-List@MadGoat.com MIME-Version: 1.0 To: MX-List@MadGoat.com CC: najman@HADASSAH.org.IL Subject: Re: trace sending computer References: <009DCC13.C7EB5CD7.14@HADASSAH.org.IL> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit "Najman Kahana. Ext:77313" wrote: > > Hello. > > I have a rather urgent, and delicate problem. > I have a user who claims that someone else is sending very nasty mail under > his name. > To do this, is child's play. > > Question: is there any way of seeing the name of the source computer ? > (The person whom I suspect would know how to change the username > in the PC mailer, but not how to change his PC's tcp name) Look at message headers, one of the "Received:" may be a help for you. It will indicate which systems mail travels through. Other headers can help too. If the mail is send in clever way, it may be impossible to identify the sender. regards Jacek ================================================================================ Archive-Date: Tue, 17 Aug 1999 18:51:54 -0700 Subject: Re: MX051 and DSN From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37b29379@news.kapsch.co.at> Date: 12 Aug 1999 11:27:21 +0100 To: MX-List@LISTS.WKU.EDU In article , Dan Wing writes: >Note you NEVER want "NOTIFY=NEVER" anyway. Such a setting would prevent >bounces such as "no such user" and "invalid domain" from being sent. >This is usually never desired and is akin to sending all of your >correspondence in the US mail without any return address. Ok. Thanks. But what is then the MX_VMSMAIL_DSN_CONTROL default setting ? Is it "NOTIFY=FAILURE" ? Which setting makes more sense ? "NOTIFY=FAILURE RETURN=HEADERS" ? >If MX is asked to do DSN processing, MX will do it. Some various >Microsoft mailers misbehave with RFC1891-compliant mailers which is I'm >sure what you're seeing. I'm not sure what I'm seeing. I only see sometimes (!) big mails going from MX "<>" to our exchange servers in our local branches and this mails are DSN mails with "successful relayed" status in the mail. So I assume, a "NOTIFY=SUCCESS,FAILURE RETURN=FULL" is the setting of this Exchange server. And I want to have "NOTIFY=FAILURE RETURN=HEADERS". And nobody (of you and me) knows how to change this and if it's even possible. Correct ? TIA -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Tue, 17 Aug 1999 20:19:51 -0700 Date: Tue, 17 Aug 1999 20:19:45 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Subject: Re: MX051 and DSN In-Reply-To: <37b29379@news.kapsch.co.at> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On 12 Aug 1999 11:27 +0100, Peter LANGSTOEGER wrote: > In article , Dan Wing writes: > >Note you NEVER want "NOTIFY=NEVER" anyway. Such a setting would prevent > >bounces such as "no such user" and "invalid domain" from being sent. > >This is usually never desired and is akin to sending all of your > >correspondence in the US mail without any return address. > > Ok. Thanks. > > But what is then the MX_VMSMAIL_DSN_CONTROL default setting ? > Is it "NOTIFY=FAILURE" ? Yes. > Which setting makes more sense ? "NOTIFY=FAILURE RETURN=HEADERS" ? Yep, that would be useful (note that it needs to be comma-separated, not space-separated). > >If MX is asked to do DSN processing, MX will do it. Some various > >Microsoft mailers misbehave with RFC1891-compliant mailers which is I'm > >sure what you're seeing. > > I'm not sure what I'm seeing. I only see sometimes (!) big mails going > from MX "<>" to our exchange servers in our local branches and this mails > are DSN mails with "successful relayed" status in the mail. Ohhhhh - the successful relay problem. Yeah, Microsoft mailers don't do this, although the RFC is quite explicit that it MUST BE done. So that's why when your Microsoft mailers are requesting this it looks like noise to your (and possibly to your users) -- Microsoft's mailers don't honor this portion of the RFC and their client mailers expect all mailers to also not honor this portion of the RFC. > So I assume, a "NOTIFY=SUCCESS,FAILURE RETURN=FULL" is the setting > of this Exchange server. And I want to have "NOTIFY=FAILURE RETURN=HEADERS". > And nobody (of you and me) knows how to change this and if it's even possible. > Correct ? Again, let me point out that any change you make to the MX_VMSMAIL_DSN_CONTROL logical will ONLY control the behavior for users that use VMSmail. From your description of your network, this is certainly _not_ the case -- your users would have to log into VMS and run the MAIL program and type stuff like "SEND" (and hit RETURN). You can change this logical all you want and the behavior you see will not change for any mail that MX is relaying. For that matter, no other product that I'm aware of, except from Microsoft, will ignore a request to handle DSNs correctly when relaying mail -- that includes PMDF and sendmail. The behavior you're seeing is described in the DSN RFCs, which are 1891 through 1894. Most specifically, see section 2.3.3 of RFC1894. -Dan Wing ================================================================================ Archive-Date: Wed, 18 Aug 1999 01:34:25 -0700 Subject: Re: MX051 and DSN From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37ba6e7d@news.kapsch.co.at> Date: 18 Aug 1999 10:27:41 +0100 To: MX-List@LISTS.WKU.EDU In article , Dan Wing writes: >On 12 Aug 1999 11:27 +0100, Peter LANGSTOEGER wrote: >> But what is then the MX_VMSMAIL_DSN_CONTROL default setting ? >> Is it "NOTIFY=FAILURE" ? > >Yes. > >> Which setting makes more sense ? "NOTIFY=FAILURE RETURN=HEADERS" ? > >Yep, that would be useful (note that it needs to be comma-separated, not >space-separated). Are you sure ? The doc says the opposite. From MX_USER_GUIDE Sect 1.7.3 $ DEFINE MX_VMSMAIL_DSN_CONTROL "NOTIFY=SUCCESS,FAILURE RETURN=HEADERS" >> I'm not sure what I'm seeing. I only see sometimes (!) big mails going >> from MX "<>" to our exchange servers in our local branches and this mails >> are DSN mails with "successful relayed" status in the mail. > >Ohhhhh - the successful relay problem. > >Yeah, Microsoft mailers don't do this, although the RFC is quite explicit >that it MUST BE done. What do the do ? What MUST BE done ? It seems that they explicitely request DSN with full headers. So what is absent ? >So that's why when your Microsoft mailers are requesting this it looks >like noise to your (and possibly to your users) -- Microsoft's mailers >don't honor this portion of the RFC and their client mailers expect >all mailers to also not honor this portion of the RFC. This is a good reason for me. But now I've to prove this (read RFC, capture a trace) and then feed it to our exchange administrators. >> So I assume, a "NOTIFY=SUCCESS,FAILURE RETURN=FULL" is the setting >> of this Exchange server. And I want to have "NOTIFY=FAILURE RETURN=HEADERS". >> And nobody (of you and me) knows how to change this and if it's even possible. >> Correct ? > >Again, let me point out that any change you make to the >MX_VMSMAIL_DSN_CONTROL logical will ONLY control the behavior for users >that use VMSmail. I did already understand this (and your last mail remembered me). > From your description of your network, this is >certainly _not_ the case -- your users would have to log into VMS >and run the MAIL program and type stuff like "SEND" (and hit RETURN). Yes, this is not the case for PC users. Most of them are now using Outlook and Exchange. Some of them (and now beginning to grow !) are M$ haters and use Netscape w/ or w/o Linux). They use MX as what it is, the company SMTP server and relay and IMAP4 and POP3 server. But I still have some (only few, I'm one of them) VMS user using VMSmail (or DECW$MAIL) and for them I should fix MX_VMSMAIL_DSN_CONTROL, too. >You can change this logical all you want and the behavior you see will >not change for any mail that MX is relaying. I know. That's why I wrote "is the setting of the exchange server". I used this logical syntax instead of a pseudo syntax used in the RFC which I still haven't read. I meant, I want to know if somebody can confirm this and perhaps even know how to change this behaviour of Exchange _if_ one can change it. >For that matter, no other product that I'm aware of, except from >Microsoft, will ignore a request to handle DSNs correctly when relaying >mail -- that includes PMDF and sendmail. > >The behavior you're seeing is described in the DSN RFCs, which are 1891 >through 1894. Most specifically, see section 2.3.3 of RFC1894. In case of M$ vs. RFC vulgo Standards, it currently seems, that M$ always wins in the head of my bosses. And I try for years to change this... Thanks -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Wed, 18 Aug 1999 08:50:31 -0700 Subject: How to reject specific To: mails ? From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37bad5b4@news.kapsch.co.at> Date: 18 Aug 1999 17:48:04 +0100 To: MX-List@LISTS.WKU.EDU I want to reject mails coming from various addresses to "To: friend@public.com". How to add a rule for this in REJMAN ? Recipient pattern is not valid with /HEADER and without /HEADER the rule is based on the envelope RCPT TO (which is a valid mailadress here of course). TIA -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Wed, 18 Aug 1999 08:52:53 -0700 Date: Wed, 18 Aug 1999 10:52:41 -0500 From: Hunter Goatley Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Message-ID: <009DCCF0.422C22F5.3@goat.process.com> Subject: RE: How to reject specific To: mails ? eplan@kapsch.net (Peter LANGSTOEGER) writes: > >I want to reject mails coming from various addresses to "To: friend@public.com". > >How to add a rule for this in REJMAN ? > >Recipient pattern is not valid with /HEADER and without /HEADER the rule >is based on the envelope RCPT TO (which is a valid mailadress here of course). > REJMAN> ADD REJECTION/HEADER "To: friend@public.com" Though I would do: REJMAN> ADD REJECTION/HEADER "To: *friend*@*public*.com" I don't know what you tried, but either of those will work. Hunter ------ Hunter Goatley, Process Software, http://www.process.com/ http://www2.wku.edu/hunter/ ================================================================================ Archive-Date: Wed, 18 Aug 1999 09:07:02 -0700 Subject: RE: How to reject specific To: mails ? From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37bad8da@news.kapsch.co.at> Date: 18 Aug 1999 18:01:30 +0100 To: MX-List@LISTS.WKU.EDU In article <009DCCF0.422C22F5.3@goat.process.com>, Hunter Goatley writes: >eplan@kapsch.net (Peter LANGSTOEGER) writes: >> >>I want to reject mails coming from various addresses to "To: friend@public.com". >> >>How to add a rule for this in REJMAN ? >> >>Recipient pattern is not valid with /HEADER and without /HEADER the rule >>is based on the envelope RCPT TO (which is a valid mailadress here of course). >> >REJMAN> ADD REJECTION/HEADER "To: friend@public.com" > >Though I would do: > >REJMAN> ADD REJECTION/HEADER "To: *friend*@*public*.com" > >I don't know what you tried, but either of those will work. Thanks for the fast answer. I already found the answer myself. It is better to read the manual than to make assumptions (which have been invalid) and then finding no explanation in HELP. ;-) My assumption was that I still have two parameters sender and recipient pattern. But instead it is only one parameter, a string, thus possibly matching all field types of a header. Sorry. -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Wed, 18 Aug 1999 09:07:09 -0700 Subject: REJMAN SHOW EXCLUSION does not work From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37bad933@news.kapsch.co.at> Date: 18 Aug 1999 18:02:59 +0100 To: MX-List@LISTS.WKU.EDU $ REJMAN SHOW ALL does show EXCLUSIONS, but $ REJMAN SHOW EXCLUSIONS doesn't show anything. Is this bug already known ? When will there be a fix ? Many TIA -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Wed, 18 Aug 1999 09:42:20 -0700 Date: Wed, 18 Aug 1999 09:42:13 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Subject: Re: MX051 and DSN In-Reply-To: <37ba6e7d@news.kapsch.co.at> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On 18 Aug 1999 10:27 +0100, Peter LANGSTOEGER wrote: > In article , Dan Wing writes: > >On 12 Aug 1999 11:27 +0100, Peter LANGSTOEGER wrote: > >> But what is then the MX_VMSMAIL_DSN_CONTROL default setting ? > >> Is it "NOTIFY=FAILURE" ? > > > >Yes. > > > >> Which setting makes more sense ? "NOTIFY=FAILURE RETURN=HEADERS" ? > > > >Yep, that would be useful (note that it needs to be comma-separated, not > >space-separated). > > Are you sure ? The doc says the opposite. From MX_USER_GUIDE Sect 1.7.3 > > $ DEFINE MX_VMSMAIL_DSN_CONTROL "NOTIFY=SUCCESS,FAILURE RETURN=HEADERS" No, I only skimmed the docs. Sorry. > >> I'm not sure what I'm seeing. I only see sometimes (!) big mails going > >> from MX "<>" to our exchange servers in our local branches and this mails > >> are DSN mails with "successful relayed" status in the mail. > > > >Ohhhhh - the successful relay problem. > > > >Yeah, Microsoft mailers don't do this, although the RFC is quite explicit > >that it MUST BE done. > > What do the do ? What MUST BE done ? > It seems that they explicitely request DSN with full headers. > So what is absent ? They should not be requesting any DSN of any kind -- then you wouldn't see the RELAY noise. > >So that's why when your Microsoft mailers are requesting this it looks > >like noise to your (and possibly to your users) -- Microsoft's mailers > >don't honor this portion of the RFC and their client mailers expect > >all mailers to also not honor this portion of the RFC. > > This is a good reason for me. But now I've to prove this (read RFC, capture > a trace) and then feed it to our exchange administrators. > > >> So I assume, a "NOTIFY=SUCCESS,FAILURE RETURN=FULL" is the setting > >> of this Exchange server. And I want to have "NOTIFY=FAILURE RETURN=HEADERS". > >> And nobody (of you and me) knows how to change this and if it's even possible. > >> Correct ? > > > >Again, let me point out that any change you make to the > >MX_VMSMAIL_DSN_CONTROL logical will ONLY control the behavior for users > >that use VMSmail. > > I did already understand this (and your last mail remembered me). Okay. > > From your description of your network, this is > >certainly _not_ the case -- your users would have to log into VMS > >and run the MAIL program and type stuff like "SEND" (and hit RETURN). > > Yes, this is not the case for PC users. Most of them are now using Outlook > and Exchange. Some of them (and now beginning to grow !) are M$ haters > and use Netscape w/ or w/o Linux). They use MX as what it is, the company > SMTP server and relay and IMAP4 and POP3 server. > > But I still have some (only few, I'm one of them) VMS user using VMSmail > (or DECW$MAIL) and for them I should fix MX_VMSMAIL_DSN_CONTROL, too. Ah, okay. > >You can change this logical all you want and the behavior you see will > >not change for any mail that MX is relaying. > > I know. That's why I wrote "is the setting of the exchange server". > I used this logical syntax instead of a pseudo syntax used in the RFC > which I still haven't read. > > I meant, I want to know if somebody can confirm this and perhaps even > know how to change this behaviour of Exchange _if_ one can change it. You might ask on microsoft.public.?.exchange if your news server has it, or check the news server at news.microsoft.com or nntp.microsoft.com or something like that. > >For that matter, no other product that I'm aware of, except from > >Microsoft, will ignore a request to handle DSNs correctly when relaying > >mail -- that includes PMDF and sendmail. > > > >The behavior you're seeing is described in the DSN RFCs, which are 1891 > >through 1894. Most specifically, see section 2.3.3 of RFC1894. > > In case of M$ vs. RFC vulgo Standards, it currently seems, that M$ always > wins in the head of my bosses. And I try for years to change this... Too true. -Dan Wing ================================================================================ Archive-Date: Wed, 18 Aug 1999 10:20:16 -0700 Subject: Re: MX051 and DSN From: eplan@kapsch.net (Peter LANGSTOEGER) Reply-To: MX-List@MadGoat.com Message-ID: <37baeaa5@news.kapsch.co.at> Date: 18 Aug 1999 19:17:25 +0100 To: MX-List@LISTS.WKU.EDU In article , Dan Wing writes: >On 18 Aug 1999 10:27 +0100, Peter LANGSTOEGER wrote: >> In article , Dan Wing writes: >> >Yeah, Microsoft mailers don't do this, although the RFC is quite explicit >> >that it MUST BE done. >> >> What do the do ? What MUST BE done ? >> It seems that they explicitely request DSN with full headers. >> So what is absent ? > >They should not be requesting any DSN of any kind -- then you wouldn't >see the RELAY noise. Of course. But why is this violating an RFC ? Still no new argument to help fighting against Exchange non-standardness ? >> I meant, I want to know if somebody can confirm this and perhaps even >> know how to change this behaviour of Exchange _if_ one can change it. > >You might ask on microsoft.public.?.exchange if your news server >has it, or check the news server at news.microsoft.com or >nntp.microsoft.com or something like that. Come on. These newsgroups are a horror and for me as a M$-Hater (see .sig) not the real place to ask. But I must admit, that I have still no other choice. And you all as MX users/managers/specialists have probably already seen Exchange servers and some of you even have them on their own site. So experience with this problem should even exist outside of the M$ circle... Still hoping... -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998 ================================================================================ Archive-Date: Wed, 18 Aug 1999 10:37:49 -0700 Date: Wed, 18 Aug 1999 12:37:38 -0500 From: Hunter Goatley Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Message-ID: <009DCCFE.EB70785D.3@goat.process.com> Subject: RE: REJMAN SHOW EXCLUSION does not work eplan@kapsch.net (Peter LANGSTOEGER) writes: > >$ REJMAN SHOW ALL does show EXCLUSIONS, but >$ REJMAN SHOW EXCLUSIONS doesn't show anything. > Use SHOW HEURISTICS to see the exclusions. Hunter ------ Hunter Goatley, Process Software, http://www.process.com/ http://www2.wku.edu/hunter/ ================================================================================ Archive-Date: Wed, 18 Aug 1999 11:55:41 -0700 Date: Wed, 18 Aug 1999 11:55:35 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: MX-List@LISTS.WKU.EDU Subject: Re: MX051 and DSN In-Reply-To: <37baeaa5@news.kapsch.co.at> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On 18 Aug 1999 19:17 +0100, Peter LANGSTOEGER wrote: > In article , Dan Wing writes: > >On 18 Aug 1999 10:27 +0100, Peter LANGSTOEGER wrote: > >> In article , Dan Wing writes: > >> >Yeah, Microsoft mailers don't do this, although the RFC is quite explicit > >> >that it MUST BE done. > >> > >> What do the do ? What MUST BE done ? > >> It seems that they explicitely request DSN with full headers. > >> So what is absent ? > > > >They should not be requesting any DSN of any kind -- then you wouldn't > >see the RELAY noise. > > Of course. But why is this violating an RFC ? The RFC says if you request a DSN but the "next hop" mailer doesn't support DSN, we have to generate a "relay" DSN. A "relay" DSN basically says: You asked for Delivery Status Notification but the next-hop mailer doesn't support Delivery Status Notification, so the best information I can give you is that I delivered the message into (what is potentially) a black hole of non-DSN-aware mailers." > Still no new argument to help fighting against Exchange non-standardness ? If Exchange finds the "relay DSN"s too noisy, Exchange shouldn't request DSNs. > >> I meant, I want to know if somebody can confirm this and perhaps even > >> know how to change this behaviour of Exchange _if_ one can change it. > > > >You might ask on microsoft.public.?.exchange if your news server > >has it, or check the news server at news.microsoft.com or > >nntp.microsoft.com or something like that. > > Come on. These newsgroups are a horror and for me as a M$-Hater (see .sig) > not the real place to ask. But I must admit, that I have still no other > choice. You won't see me there, either! > And you all as MX users/managers/specialists have probably already seen > Exchange servers and some of you even have them on their own site. So > experience with this problem should even exist outside of the M$ circle... > > Still hoping... I recall some other posts on this subject to mx-list -- I'm surprised no one else has spoken up. -d ================================================================================ Archive-Date: Wed, 18 Aug 1999 11:55:50 -0700 Date: Wed, 18 Aug 1999 11:55:35 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: MX-List@LISTS.WKU.EDU Subject: Re: MX051 and DSN In-Reply-To: <37baeaa5@news.kapsch.co.at> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On 18 Aug 1999 19:17 +0100, Peter LANGSTOEGER wrote: > In article , Dan Wing writes: > >On 18 Aug 1999 10:27 +0100, Peter LANGSTOEGER wrote: > >> In article , Dan Wing writes: > >> >Yeah, Microsoft mailers don't do this, although the RFC is quite explicit > >> >that it MUST BE done. > >> > >> What do the do ? What MUST BE done ? > >> It seems that they explicitely request DSN with full headers. > >> So what is absent ? > > > >They should not be requesting any DSN of any kind -- then you wouldn't > >see the RELAY noise. > > Of course. But why is this violating an RFC ? The RFC says if you request a DSN but the "next hop" mailer doesn't support DSN, we have to generate a "relay" DSN. A "relay" DSN basically says: You asked for Delivery Status Notification but the next-hop mailer doesn't support Delivery Status Notification, so the best information I can give you is that I delivered the message into (what is potentially) a black hole of non-DSN-aware mailers." > Still no new argument to help fighting against Exchange non-standardness ? If Exchange finds the "relay DSN"s too noisy, Exchange shouldn't request DSNs. > >> I meant, I want to know if somebody can confirm this and perhaps even > >> know how to change this behaviour of Exchange _if_ one can change it. > > > >You might ask on microsoft.public.?.exchange if your news server > >has it, or check the news server at news.microsoft.com or > >nntp.microsoft.com or something like that. > > Come on. These newsgroups are a horror and for me as a M$-Hater (see .sig) > not the real place to ask. But I must admit, that I have still no other > choice. You won't see me there, either! > And you all as MX users/managers/specialists have probably already seen > Exchange servers and some of you even have them on their own site. So > experience with this problem should even exist outside of the M$ circle... > > Still hoping... I recall some other posts on this subject to mx-list -- I'm surprised no one else has spoken up. -d ================================================================================ Archive-Date: Wed, 18 Aug 1999 14:08:38 -0700 Date: Wed, 18 Aug 1999 16:12:59 -0500 From: system@niuhep.physics.niu.edu Reply-To: MX-List@MadGoat.com To: MX-LIST@MADGOAT.COM Message-ID: <009DCD1D.00D16D24.30@niuhep.physics.niu.edu> Subject: locally changed from lines Hello, A user wants to change the from line on netscape so that mail he sends from here looks like it comes from his more usual address. from: joe@netsector.com I made the following changes and did a reset but he still gets a message saying relaying not allowed. Local domain patterns for SMTP relay checks: NETSECTOR.COM Inside IP networks/hosts: 216.157.1.161 netmask 255.255.255.255 (Relay allowed) I am guessing that this is running afoul of a totally different anti-relay check. Any advice? Robert ================================================================================ Archive-Date: Wed, 18 Aug 1999 16:39:29 -0700 Date: Wed, 18 Aug 1999 16:39:16 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Subject: Re: locally changed from lines In-Reply-To: <009DCD1D.00D16D24.30@niuhep.physics.niu.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Wed, 18 Aug 1999 16:12 -0500, system@niuhep.physics.niu.edu wrote: > A user wants to change the from line on netscape so that mail he sends from > here looks like it comes from his more usual address. > > from: joe@netsector.com > > I made the following changes and did a reset but he still gets a message > saying relaying not allowed. > > Local domain patterns for SMTP relay checks: > NETSECTOR.COM > > Inside IP networks/hosts: > 216.157.1.161 netmask 255.255.255.255 (Relay allowed) > > I am guessing that this is running afoul of a totally different > anti-relay check. > > Any advice? Get a TCPDUMP of the user trying to send mail to make sure it is really MX that is generating the message. I assume you have reset the MX agents after the config change? -dan $ MULTINET TCPDUMP/HEX/SNAP=1514 PORT 25 AND HOST ip_address_of_pc ================================================================================ Archive-Date: Thu, 19 Aug 1999 08:09:02 -0700 Date: Thu, 19 Aug 1999 09:10:23 -0500 From: system@niuhep.physics.niu.edu Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Message-ID: <009DCDAB.21F00810.3@niuhep.physics.niu.edu> Subject: Re: locally changed from lines From: Dan Wing :Get a TCPDUMP of the user trying to send mail to make sure it is really MX :that is generating the message. He is using the VMS netscape, the message is coming directly from netscape. He was able to do this before. Hmmm, I just tried it and succeeded. Maybe he is trying to use his usual server as the SMTP server? I guess I'll have to look at his configuration more carefully. Robert ================================================================================ Archive-Date: Thu, 19 Aug 1999 08:40:36 -0700 Date: Thu, 19 Aug 1999 08:40:29 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com Subject: Re: locally changed from lines In-Reply-To: <009DCDAB.21F00810.3@niuhep.physics.niu.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Thu, 19 Aug 1999 09:10 -0500, system@niuhep.physics.niu.edu wrote: > From: Dan Wing > > :Get a TCPDUMP of the user trying to send mail to make sure it is really MX > :that is generating the message. > > He is using the VMS netscape, the message is coming directly from > netscape. He was able to do this before. It still connects to port 25 to send mail but may connect using the lo0 interface instead of se0: $ multinet tcpdump/hex/snap=1514/interface=lo0 port 25 -d ================================================================================ Archive-Date: Fri, 20 Aug 1999 14:28:03 -0700 Date: Fri, 20 Aug 1999 16:57:41 -0500 From: "Jonathan E. Hardis" Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: system@niuhep.physics.niu.edu, hardis@garnet.nist.gov Message-ID: <009DCEB5.944930C0.1@garnet.nist.gov> Subject: RE: locally changed from lines > Inside IP networks/hosts: > 216.157.1.161 netmask 255.255.255.255 (Relay allowed) Check the netmask. Should it be 255.255.255.0 ? - Jonathan ================================================================================ Archive-Date: Tue, 24 Aug 1999 13:15:37 -0700 Message-ID: <4.1.19990824151041.00ac8220@nicanor.acu.edu> Date: Tue, 24 Aug 1999 15:15:30 -0500 To: mx-list@wku.edu From: Tom Dolan Reply-To: MX-List@MadGoat.com Subject: Urgent - smtp problem MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" For some reason my MX SMTP processes quit delivering email to nicanor.acu.edu - about 450 messages backed up in the que waiting to be delivered to that machine. The MCP show que /full error message says Last error: %MX-W-HOSTLOCK, target host is temporarily locked against conn ection attempts However I can telnet nicanor.acu.edu /port=25 with no problem. What do I do now? Tom Dolan Dolan@Bible.acu.edu 242 Bible Building Systems Manager ACU Box 29454 College of Biblical Studies Abilene, TX 79699 Abilene Christian University 915.674.3706 http://dolan.acu.edu ================================================================================ Archive-Date: Tue, 24 Aug 1999 13:54:18 -0700 Message-ID: <001001beee73$4e7c95f0$14c50718@lvrmr1.sfba.home.com> From: "Esa Laitinen" Reply-To: MX-List@MadGoat.com To: CC: References: <4.1.19990824151041.00ac8220@nicanor.acu.edu> Subject: Re: Urgent - smtp problem Date: Tue, 24 Aug 1999 13:57:41 -0700 > What do I do now? There has been a glitch that has caused MX to think that the server is unreachable. It should try again later. You can delete MX_SMTP_LOCK_DIR: directory contents so the mail will start flowing, each entry when the retry interval is up, I guess. You can ready the entries to start mail moving faster. Yours, Esa Laitinen ================================================================================ Archive-Date: Tue, 24 Aug 1999 15:56:08 -0700 Date: Tue, 24 Aug 1999 15:56:02 -0700 (PDT) From: Dan Wing Reply-To: MX-List@MadGoat.com To: MX-List@MadGoat.com CC: Dolan@Bible.acu.edu Subject: Re: Urgent - smtp problem In-Reply-To: <001001beee73$4e7c95f0$14c50718@lvrmr1.sfba.home.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Tue, 24 Aug 1999 13:57 -0700, Esa Laitinen wrote: > > What do I do now? > > There has been a glitch that has caused MX to think that the server is > unreachable. It should try again later. It will retry every 30 minutes, with or without the lockfile. The lockfile merely prevents each of those 450 entries from being retried individually. -d ================================================================================ Archive-Date: Wed, 25 Aug 1999 05:28:32 -0700 Message-ID: <06f801beeef5$5ae04d00$e0680e26@hmiller> Reply-To: MX-List@MadGoat.com From: "Henry W. Miller" To: , References: <4.1.19990824151041.00ac8220@nicanor.acu.edu> Subject: Re: Urgent - smtp problem Date: Wed, 25 Aug 1999 05:28:34 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit ----- Original Message ----- From: Tom Dolan To: Sent: Tuesday, August 24, 1999 13:15 PM Subject: Urgent - smtp problem > For some reason my MX SMTP processes quit delivering email to > nicanor.acu.edu - about 450 messages backed up in the que waiting > to be delivered to that machine. > > The MCP show que /full error message says > > Last error: %MX-W-HOSTLOCK, target host is temporarily locked > against conn > ection attempts > > However I can telnet nicanor.acu.edu /port=25 with no problem. > > What do I do now? > > Tom Dolan Dolan@Bible.acu.edu 242 Bible Building > Systems Manager ACU Box 29454 > College of Biblical Studies Abilene, TX 79699 > Abilene Christian University 915.674.3706 > http://dolan.acu.edu > Tom, As others have mentioned, try deleting the MX lock file for this host. However, I've just tried several experiments with connecting to nicanor.acu.edu, without success. I have been able to connect to their backup MX site, noah.acu.edu. So, maybe the problem lies mainly at their site? -HWM ================================================================================ Archive-Date: Tue, 31 Aug 1999 14:16:49 -0700 Date: Tue, 31 Aug 1999 15:15:12 -0600 (CST) From: "A. Mahendra Rajah, Systems Manager" Reply-To: MX-List@MadGoat.com Subject: Setting up anti-relay in MX 5.0 To: MX-List@MadGoat.com CC: A.Mahendra.Rajah@Meena.CC.URegina.CA Message-ID: <01JFF71W6JCU8Y6CJB@meena.cc.uregina.ca> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII We need to disable unauthorized SMTP relaying through our v5.0 MX server. We thought we could simply set MCP SET SMTP/NORELAY_ALLOWED But, after reading the description of this in section 7.4 in the MX Management Guide, we are not sure if this will work for us. Here is the scenerio: Let us call our server machine A. a) Nobody logs direclty into A nor use any mail clients connecting to A to read mail messages. b) All those who have accounts on A have mail forwarding set to their work/home addresses which they update whenever they change jobs or ISPs. c) Most of these people have subscribed to mailing lists using their email addresses on A. d) There are several mailing lists defined on A using MXserver. Members to these lists may or may not have accounts on A. So, if we did the MCP SET SMTP/NORELAY_ALLOWED command, what will happen to mail meant for those in b), c) and d)? Are there better ways to achieve this? The only method that I can find is to MCP DEFINE LOCAL_DOMAIN for every valid recipient's address in b) and d). Ouch! All help will be appreciated. -- mahendra ......................................................................... A. Mahendra Rajah Email: Mahendra@Meena.CC.URegina.CA VMS & DUNIX Systems Manager Phone: (306) 585-4496 Dept. of Computing Services FAX: (306) 585-5060 University of Regina, Regina, Sask., S4S 0A2 Canada. .........................................................................