Re: qmail remote delivery logic

1999-11-16 Thread Fred Lindberg
On Mon, 15 Nov 1999 22:31:52 -0600, Bruce Guenter wrote: 2a. if -l: compare host name of envelope recipients. If same as local host name, deliver locally with qmail-local. 2b. deliver multi-recipient message remotely (concurrent with 2a). If -l: remove all recipients with host part

Re: qmail remote delivery logic

1999-11-15 Thread Bruce Guenter
On Sun, Nov 14, 1999 at 08:50:33AM +, Frederik Lindberg wrote: Was looking for something optimized to low bandwidth settings. *slaps self* I'm pretty forgetful some times. So let me ask some more questions: 2a. if -l: compare host name of envelope recipients. If same as local host

Re: qmail remote delivery logic

1999-11-14 Thread Frederik Lindberg
On Sat, 13 Nov 1999, Bruce Guenter wrote: 1. message queued. 2a. if -l: compare host name of envelope recipients. If same as local host name, deliver locally with qmail-local. 2b. deliver multi-recipient message remotely (concurrent with 2a). If -l: remove all recipients with host part

Re: qmail remote delivery logic

1999-11-13 Thread Bruce Guenter
On Mon, Nov 08, 1999 at 04:50:32AM +, Frederik Lindberg wrote: Hard, but not impossible. How would you envision such hooks? How about a simplified scheme: All mail with recipient host matching the local host is passed to an external program for delivery. The external program takes

Re: qmail remote delivery logic

1999-11-11 Thread Dave Sill
Sam [EMAIL PROTECTED] wrote: ... Qmail will also do pretty well when the mail volume is low, although there are certain pathological situations where Qmail will fail miserably with low mail volume. What situations are you referring to? Also, Qmail will do poorly in the extreme upper end of the

Re: qmail remote delivery logic

1999-11-10 Thread Russell Nelson
Frederik Lindberg writes: Qmail is fastest under well-connected conditions when the recipients per host are close to 1 (most "normal traffic"). Still, it would be nice to improve it in more marginal situations as well. QMTP to smarthosts and other qmail hosts would go a long way, although

Re: qmail remote delivery logic

1999-11-09 Thread Steve Vertigan
Dave Sill wrote: Say you send a message to a list of 10,000 addresses using sendmail. What's the first thing it does? It looks up the MX for each recipient so it can sort by MX and minimize the number of connections. Why is that? Lets say you have to deliver to [EMAIL PROTECTED], [EMAIL

Re: qmail remote delivery logic

1999-11-09 Thread Dave Sill
[EMAIL PROTECTED] wrote: On Mon, 8 Nov 1999, Dave Sill wrote: Say you send a message to a list of 10,000 addresses using sendmail. What's the first thing it does? It looks up the MX for each recipient so it can sort by MX and minimize the number of connections. I doubt very much that that's

Re: qmail remote delivery logic

1999-11-09 Thread Sam
David Dyer-Bennet writes: Sam [EMAIL PROTECTED] writes on 8 November 1999 at 17:39:02 -0500 Pard'n me, but how does an additional DNS lookup for every recipient end up reducing the overall amount of DNS traffic? qmail does fewer DNS lookups than sendmail, and that should be Not

Re: qmail remote delivery logic

1999-11-09 Thread Sam
Stefan Paletta writes: Sam wrote/schrieb/scribsit: different domains will result in only 5,000 DNS queries. Meanwhile, each instance of qmail-remote should diligently issue a DNS query - for a grand sum of 10,000 queries overall. When we're talking about lists of that size, you will

Re: qmail remote delivery logic

1999-11-08 Thread Dave Sill
"Jim B" [EMAIL PROTECTED] wrote: Ummm.. yeah thanks. But I want to know *why* it's faster. I know there's a doc that explains 3 different methods, this being one of them... and it shows situations why one may be preferable over the other. Do you know what doc I'm talking about?

Re: qmail remote delivery logic

1999-11-08 Thread Frank Tegtmeyer
In a pathological case, qmail can use a lot more network bandwidth because of the duplication of messages going to the same system. In practice this is rarely a serious problem. Taking into account the *decreased* DNS traffic, it's even more rarely a problem. It depends heavily on the

Re: qmail remote delivery logic

1999-11-08 Thread Pavel Kankovsky
On Mon, 8 Nov 1999, David Dyer-Bennet wrote: To start with, you have to do a HUGE number of extra DNS lookups to determine what the recipient systems ARE. As opposed to the "one SMTP transaction per remote recipient" strategy that is able to deliver messages without the need to figure out

Re: qmail remote delivery logic

1999-11-08 Thread Dave Sill
Pavel Kankovsky [EMAIL PROTECTED] wrote: On Mon, 8 Nov 1999, David Dyer-Bennet wrote: To start with, you have to do a HUGE number of extra DNS lookups to determine what the recipient systems ARE. As opposed to the "one SMTP transaction per remote recipient" strategy that is able to deliver

Re: qmail remote delivery logic

1999-11-08 Thread Pavel Kankovsky
On Mon, 8 Nov 1999, Dave Sill wrote: Pavel Kankovsky [EMAIL PROTECTED] wrote: On Mon, 8 Nov 1999, David Dyer-Bennet wrote: To start with, you have to do a HUGE number of extra DNS lookups to determine what the recipient systems ARE. As opposed to the "one SMTP transaction per remote

Re: qmail remote delivery logic

1999-11-08 Thread Dave Sill
Pavel Kankovsky [EMAIL PROTECTED] wrote: Indeed. But are there any EXTRA lookups done? ("Extra" is the keyword here...read original DDB's text again.) The answer is: NO unless the implementation is incredibly stupid. Like sendmail? :-) Sendmail is notorious for unnecessary DNS lookups. I

Re: qmail remote delivery logic

1999-11-08 Thread Dave Sill
"Frank Tegtmeyer" [EMAIL PROTECTED] wrote: It's amazing how qmail haters (here in Germany) always reduced the discussion about qmail to this special case - it may be bad discussion style but I also think that there is more need to support this type of setup than the "normal" qmail

Re: qmail remote delivery logic

1999-11-08 Thread David L. Nicol
We have the source; let's fix it. What the people with the problem are asking for appears to be for qmail to not split up identical mails intended for multiple recipients at identical hosts. These are real problems and poo-pooing them as degenerate cases or something produces nothing. In

Re: qmail remote delivery logic

1999-11-08 Thread John White
I understand the motivation David, I really do. But you don't seem to understand who qmail works. On Mon, Nov 08, 1999 at 09:12:48PM +, David L. Nicol wrote: What the people with the problem are asking for appears to be for qmail to not split up identical mails intended for multiple

Re: qmail remote delivery logic

1999-11-08 Thread John White
On Mon, Nov 08, 1999 at 11:24:23PM +, David L. Nicol wrote: John White flamed forth: man pages indicate ... that qmail-remote "sends the message to one or more recipients at a remote host." Which means that it still hasn't been split up when qmail-remote gets it, and that

qmail remote delivery logic

1999-11-07 Thread Jim B
Can someone point me at the location of a document that explains why qmail would deliver, for example, a msg to 5 recipients at the same remote domain with 5 individual smtp connections instead of one smtp connection and multiple RCPT TOs? I just spotted it the other day and meant to go back

Re: qmail remote delivery logic

1999-11-07 Thread Russell Nelson
Jim B writes: Can someone point me at the location of a document that explains why qmail would deliver, for example, a msg to 5 recipients at the same remote domain with 5 individual smtp connections instead of one smtp connection and multiple RCPT TOs? Because it's faster. -- -russ

Re: qmail remote delivery logic

1999-11-07 Thread Sam
Russell Nelson writes: Jim B writes: Can someone point me at the location of a document that explains why qmail would deliver, for example, a msg to 5 recipients at the same remote domain with 5 individual smtp connections instead of one smtp connection and multiple RCPT TOs?

Re: qmail remote delivery logic

1999-11-07 Thread Jeff Hayward
On Sun, 7 Nov 1999, Russell Nelson wrote: Because it's faster. For the average message... :-) -- Jeff

Re: qmail remote delivery logic

1999-11-07 Thread Eric Dahnke
Could someone explain how qmail manages to be faster for average msgs. I can't see how it would be. - Eric Jeff Hayward escribió: On Sun, 7 Nov 1999, Russell Nelson wrote: Because it's faster. For the average message... :-) -- Jeff

Re: qmail remote delivery logic

1999-11-07 Thread Russell Nelson
Sam writes: Russell Nelson writes: Jim B writes: Can someone point me at the location of a document that explains why qmail would deliver, for example, a msg to 5 recipients at the same remote domain with 5 individual smtp connections instead of one smtp connection and

Re: qmail remote delivery logic

1999-11-07 Thread Russell Nelson
Eric Dahnke writes: Could someone explain how qmail manages to be faster for average msgs. I can't see how it would be. The most-oft used MTA fiddles with hostnames while the DNS burns. -- -russ nelson [EMAIL PROTECTED] http://russnelson.com Crynwr sells support for free software |

Re: qmail remote delivery logic

1999-11-07 Thread Jim B
Nelson [EMAIL PROTECTED] To: Jim B [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Sunday, November 07, 1999 2:10 PM Subject: Re: qmail remote delivery logic Jim B writes: Can someone point me at the location of a document that explains why qmail would deliver, for example, a msg to 5 recipients

Re: qmail remote delivery logic

1999-11-07 Thread Sam
On Sun, 7 Nov 1999, Russell Nelson wrote: Sam writes: Russell Nelson writes: Jim B writes: Can someone point me at the location of a document that explains why qmail would deliver, for example, a msg to 5 recipients at the same remote domain with 5 individual smtp

Re: qmail remote delivery logic

1999-11-07 Thread Russell Nelson
Sam writes: Yes. Even off a T1, there's a measurable difference between ~10 MB and ~1 MB worth of traffic. And?? Don't hold us in suspense. What was the difference in delivery times? -- -russ nelson [EMAIL PROTECTED] http://russnelson.com Crynwr sells support for free software | PGPok

Re: qmail remote delivery logic

1999-11-07 Thread Jason Haar
On Sun, Nov 07, 1999 at 04:21:20PM -0500, Russell Nelson wrote: Sam writes: Yes. Even off a T1, there's a measurable difference between ~10 MB and ~1 MB worth of traffic. And?? Don't hold us in suspense. What was the difference in delivery times? I was going to keep quiet but how

Re: qmail remote delivery logic

1999-11-07 Thread Andy Bradford
Thus said Jason Haar on Mon, 08 Nov 1999 10:36:38 +1300: We have a 64Kb Frame Relay link with burst to 128Kb. We have users here sending their current favourite 3 Mb MP3 file to 30 friends - effectively taking our Internet link offline for the next several hours. Qmail being the great

Re: qmail remote delivery logic

1999-11-07 Thread James J. Lippard
On Sun, 07 Nov 1999 15:11:29 -0700 in Andy Bradford [EMAIL PROTECTED] wrote: Thus said Jason Haar on Mon, 08 Nov 1999 10:36:38 +1300: We have a 64Kb Frame Relay link with burst to 128Kb. We have users here sending their current favourite 3 Mb MP3 file to 30 friends - effectively taking

Re: qmail remote delivery logic

1999-11-07 Thread Fred Lindberg
On Mon, 8 Nov 1999 10:36:38 +1300, Jason Haar wrote: We have a 64Kb Frame Relay link with burst to 128Kb. We have users here sending their current favourite 3 Mb MP3 file to 30 friends - effectively [...] Fact: Sendmail would have used less bandwidth in this _specific_ situation. In general -

Re: qmail remote delivery logic

1999-11-07 Thread Andy Bradford
Thus said "James J. Lippard" on 07 Nov 1999 17:12:54 MST: Times are changing. Unified messaging is coming. Email, voice mail, faxes, video mail, all will be the same thing. "User education" will not be the answer--building the appropriate user interfaces and designing the appropriate

Re: qmail remote delivery logic

1999-11-07 Thread Jason Haar
On Sun, Nov 07, 1999 at 05:22:43PM -0700, Andy Bradford wrote: Thus said "James J. Lippard" on 07 Nov 1999 17:12:54 MST: Times are changing. Unified messaging is coming. Email, voice mail, faxes, video mail, all will be the same thing. "User education" will not be the answer--building

Re: qmail remote delivery logic

1999-11-07 Thread Bruce Guenter
On Sun, Nov 07, 1999 at 06:20:36PM -0600, Fred Lindberg wrote: You could also look into QMQP over your link. [...] AFAIK, Bruce Guenter's nullmailer does this and can use QMTP to the smarthosts. nullmailer can indeed use QMTP to smarthosts that support it, but it explicitly has no support for

Re: qmail remote delivery logic

1999-11-07 Thread Frederik Lindberg
On Sun, 7 Nov 1999, Bruce Guenter wrote: nullmailer can indeed use QMTP to smarthosts that support it, but it explicitly has no support for local delivery. You could set up qmail and nullmailer concurently and have qmail deliver all remote mail into a virtual domain that calls up nullmailer

Re: qmail remote delivery logic

1999-11-07 Thread Bruce Guenter
On Mon, Nov 08, 1999 at 04:13:25AM +, Frederik Lindberg wrote: nullmailer can indeed use QMTP to smarthosts that support it, but it explicitly has no support for local delivery. You could set up qmail and nullmailer concurently and have qmail deliver all remote mail into a virtual

Re: qmail remote delivery logic

1999-11-07 Thread Frederik Lindberg
On Sun, 7 Nov 1999, Bruce Guenter wrote: Hard, but not impossible. How would you envision such hooks? Certainly, I am not going to add support to nullmailer to actually do the delivery, but giving it a way to call an external program that could is a reasonable option. How about a