From brian at enchanter.net Wed Jan 1 14:11:34 2003 From: brian at enchanter.net (Brian Kendig) Date: Wed, 1 Jan 2003 09:11:34 -0500 Subject: [SA-exim] Problem with the logging patch of the other day Message-ID: Marc -- there's a slight problem with the patch you did the other day to avoid writing that 'savemail condition expand returned false' line to the log every time. The problem is that, now, all incoming mail gets logged, even when it's not supposed to be. Here's your code at line 158 of sa-exim.c: if (SAEximDebug > 1 && ( expand[0] == 0 || (expand[0] == '0' && expand[1] == 0))) { log_write(0, LOG_MAIN, "SA: savemail condition expanded to false, not saving message to disk"); return 0; } I think you want something like this instead: if (expand[0] == 0 || (expand[0] == '0' && expand[1] == 0)) { if (SAEximDebug > 1) { log_write(0, LOG_MAIN, "SA: savemail condition expanded to false, not saving message to disk"); } return 0; } ____ |\/| Brian Kendig Set your priorities right. \ /\ / ..__. brian at enchanter net No one ever said on his \/ \__\ _/ http://www.enchanter.net/ death bed, "Gee, if I'd \__ __ \_ Be insatiably curious. only spent more time at \____\___\ Ask "why" a lot. the office." From brian at enchanter.net Wed Jan 1 14:20:21 2003 From: brian at enchanter.net (Brian Kendig) Date: Wed, 1 Jan 2003 09:20:21 -0500 Subject: [SA-exim] Who's using teergrubing? Message-ID: <296BBE6A-1D94-11D7-A39F-003065546CF4@enchanter.net> I'm curious -- how many people out there are running teergrubes? And what do you have your teergrube delay set to? I'm using eight hours for mine, and so far I haven't caught more than about twenty spammers at any one time. Also, is there any way to keep the size of the spawned exim processes down while they're teergrubing spammers? Right now each of my exim processes is about 600KB resident with a little over 2MB swapped to disk. So twenty spammers tie up about twelve megabytes of RAM and over 40MB of disk, and if I ever manage to snare two or three times as many spammers, it might start making a dent in my server's resources. Since the teergrube process is only echoing text over the connection and writing to the exim log, is there any way to pare it down significantly? ____ |\/| Brian Kendig Set your priorities right. \ /\ / ..__. brian at enchanter net No one ever said on his \/ \__\ _/ http://www.enchanter.net/ death bed, "Gee, if I'd \__ __ \_ Be insatiably curious. only spent more time at \____\___\ Ask "why" a lot. the office." From marc at merlins.org Thu Jan 2 16:44:05 2003 From: marc at merlins.org (Marc MERLIN) Date: Thu, 2 Jan 2003 17:44:05 +0100 Subject: [SA-exim] Problem with the logging patch of the other day In-Reply-To: References: Message-ID: <20030102164404.GA9640@merlins.org> On Wed, Jan 01, 2003 at 09:11:34AM -0500, Brian Kendig wrote: > Marc -- there's a slight problem with the patch you did the other day > to avoid writing that 'savemail condition expand returned false' line > to the log every time. The problem is that, now, all incoming mail > gets logged, even when it's not supposed to be. Yes, you are absolutely right, that's what I get for coding while being jetlagged, and a bit sick due to too much food :) > I think you want something like this instead: > > if (expand[0] == 0 || (expand[0] == '0' && expand[1] == 0)) > { > if (SAEximDebug > 1) > { > log_write(0, LOG_MAIN, "SA: savemail condition expanded to > false, not saving message to disk"); > } > return 0; > } Absolutely, I just checked that in. Thanks for reviewing my cvs patch Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems & security .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ | Finger marc_f@merlins.org for PGP key From marc at merlins.org Fri Jan 3 11:13:30 2003 From: marc at merlins.org (Marc MERLIN) Date: Fri, 3 Jan 2003 12:13:30 +0100 Subject: [SA-exim] Who's using teergrubing? In-Reply-To: <296BBE6A-1D94-11D7-A39F-003065546CF4@enchanter.net> References: <296BBE6A-1D94-11D7-A39F-003065546CF4@enchanter.net> Message-ID: <20030103111329.GA4456@merlins.org> On Wed, Jan 01, 2003 at 09:20:21AM -0500, Brian Kendig wrote: > I'm curious -- how many people out there are running teergrubes? And I am :-) > what do you have your teergrube delay set to? I'm using eight hours 10mn. I find that most if not all MTAs have code to make sure they don't hang around forever. Many don't stay longer than a few minutes, for that matter, many spammers don't stay longer than 30 secs > Also, is there any way to keep the size of the spawned exim processes > down while they're teergrubing spammers? Right now each of my exim > processes is about 600KB resident with a little over 2MB swapped to > disk. So twenty spammers tie up about twelve megabytes of RAM and over No, I don't agree. Almost all that memory should be shared. I don't know how much RAM is malloced by exim, but only that RAM should be duplicated by each process. Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems & security .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ | Finger marc_f@merlins.org for PGP key From marc at merlins.org Fri Jan 3 12:18:13 2003 From: marc at merlins.org (Marc MERLIN) Date: Fri, 3 Jan 2003 13:18:13 +0100 Subject: [SA-exim] Re: SAmaxrcptlistlength should have a different default In-Reply-To: <1E40352B-1E8B-11D7-8C7B-003065546CF4@enchanter.net> References: <20030102164404.GA9640@merlins.org> <1E40352B-1E8B-11D7-8C7B-003065546CF4@enchanter.net> Message-ID: <20030103121812.GD4456@merlins.org> On Thu, Jan 02, 2003 at 02:48:08PM -0500, Brian Kendig wrote: > I'd like to suggest that this should be changed slightly... I don't > think there's any reason why anyone would want to limit the length of > rcpt lists to less than what Exim supports (is there?). So why not > default to the maximum length, and use a 'SAhidercptlist: 1' setting > instead if an admin would like to hide the rcpt list for privacy > reasons? Mmmh, I'm not sure I agree on that one. 1) People may want to limit the header to let's say 160 chars or so. A 8K header can be huge and maybe even crash stupid MUAs or relaying MTAs 2) Yes, I changed the default (and documented it as changed) as people pointed out to me that defaulting to having that header on was probably not a good idea, even if I documented it as such. So now, you have to enable the header (I think you just need to comment out the entry in the conf file) to get it. Do you think it's unreasonable? Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems & security .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ | Finger marc_f@merlins.org for PGP key From brian at enchanter.net Fri Jan 3 14:11:41 2003 From: brian at enchanter.net (Brian Kendig) Date: Fri, 3 Jan 2003 09:11:41 -0500 Subject: [SA-exim] Re: SAmaxrcptlistlength should have a different default In-Reply-To: <20030103121812.GD4456@merlins.org> Message-ID: <483A0B4A-1F25-11D7-8EF8-000A27952E04@enchanter.net> Marc MERLIN wrote: > Do you think it's unreasonable? Naaw, what you said makes sense. I use my mail server for my custom domain with me as the only user; I forget sometimes that other people have different needs for it. ;-) I guess it makes sense to leave that header off by default and allow it to be set to something reasonably small. I was just caught by surprise because I didn't notice the documentation of this change (I didn't look hard enough, but other people upgrading might be surprised by it too, so you might want to hilight it as much as you can!). From brian at enchanter.net Fri Jan 3 14:42:31 2003 From: brian at enchanter.net (Brian Kendig) Date: Fri, 3 Jan 2003 09:42:31 -0500 Subject: [SA-exim] Who's using teergrubing? In-Reply-To: <20030103111329.GA4456@merlins.org> Message-ID: <96B84E34-1F29-11D7-8EF8-000A27952E04@enchanter.net> Marc MERLIN wrote: >> what do you have your teergrube delay set to? I'm using eight hours > > 10mn. I find that most if not all MTAs have code to make sure they > don't > hang around forever. Many don't stay longer than a few minutes, for > that > matter, many spammers don't stay longer than 30 secs Actually, I'm finding otherwise. I arbitrarily grabbed my stats for a week ago, which happens to be December 27, to have a look at 'em... On that day, my server identified 119 messages as having a SpamAssassin score of 30.0 or above, which is the score I use to start teergrubing them. 52 spammers waited patiently for all eight hours of my teergrube delay. 33 spammers hung around for between 15 minutes and 8 hours, but left before the full delay finished. The remaining 34 spammers dropped before 20 seconds. I feel happy to have been able to put a dent in the operations of 71% of the spammers who dropped by that day. :) > Almost all that memory should be shared. Whups, I forgot about that... thanks for reminding me! Sometimes I wish I could do more to spammers than simply holding 'em in place. Do mailers typically log all their SMTP traffic? Would it make a difference if, instead of sending a spammer one line of text every ten seconds, I sent him the entire contents of /usr/dict/words every ten seconds? Are there any known holes in spamware that my server could exploit to cause it to crash? From marc at merlins.org Fri Jan 3 23:45:56 2003 From: marc at merlins.org (Marc MERLIN) Date: Sat, 4 Jan 2003 00:45:56 +0100 Subject: [SA-exim] Re: SAmaxrcptlistlength should have a different default In-Reply-To: <483A0B4A-1F25-11D7-8EF8-000A27952E04@enchanter.net> References: <20030103121812.GD4456@merlins.org> <483A0B4A-1F25-11D7-8EF8-000A27952E04@enchanter.net> Message-ID: <20030103234556.GF6239@merlins.org> On Fri, Jan 03, 2003 at 09:11:41AM -0500, Brian Kendig wrote: > I was just caught by surprise because I didn't notice the documentation > of this change (I didn't look hard enough, but other people upgrading > might be surprised by it too, so you might want to hilight it as much > as you can!). It will be in the changelog once I release the next version, as you mention this is a change that needs to be mentionned more than once. If you use CVS, you should diff the README before you update your tree :-) Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems & security .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ | Finger marc_f@merlins.org for PGP key From plonka at doit.wisc.edu Tue Jan 7 19:27:08 2003 From: plonka at doit.wisc.edu (Dave Plonka) Date: Tue, 7 Jan 2003 13:27:08 -0600 Subject: [SA-exim] Debian issue w/exim4_i386.deb, depends on libc6 2.2.5-14.3 Message-ID: <20030107132708.B19086@doit.wisc.edu> SA-Exim users, I'm using Marc's (sa-)exim4 Debian package on my machines running the stable "woody" distribution. The problem is that the current stable libc6 is version 2.2.5-11.2, but the "exim4_i386.deb" here: http://marc.merlins.org/linux/exim/files/exim4_i386.deb wants 2.2.5-14.3, which happens to be the current version from "testing" distribution. As an q&d workaround, I installed it like this: dpkg --install --ignore-depends=libc6 exim4_i386.deb and sa-exim works fine, but now dselect is very unhappy. It won't let me do anything else (unless I switch from "exim4" back to "exim") because the dependency rules were broken. Is there an easy way for me to rebuild the package to just depend on the stable libc6 - or is this something you could fix, Marc? It would make the package more useful to those of us trying to stay with the stable code. If this is a misguided question, other Debian users please suggest to me what my best strategy is. Should I have just built from source and stayed out of the way of the Debian packages? I'm afraid to switch to "testing" since I've always just run the "stable" distribution. If I change to "testing" in "source.list", dselect will want to upgrade all packages to the "testing" version, right? Ugh. Thanks, Dave -- plonka@doit.wisc.edu http://net.doit.wisc.edu/~plonka ARS:N9HZF Madison, WI From marc at merlins.org Tue Jan 7 19:38:14 2003 From: marc at merlins.org (Marc MERLIN) Date: Tue, 7 Jan 2003 11:38:14 -0800 Subject: [SA-exim] Debian issue w/exim4_i386.deb, depends on libc6 2.2.5-14.3 In-Reply-To: <20030107132708.B19086@doit.wisc.edu> References: <20030107132708.B19086@doit.wisc.edu> Message-ID: <20030107193814.GK4197@merlins.org> On Tue, Jan 07, 2003 at 01:27:08PM -0600, Dave Plonka wrote: > wants 2.2.5-14.3, which happens to be the current version from "testing" > distribution. Ah, yeah, darn. If someone who knows debian can tell me how to force an older version of the libc when I build the package, I'll be happy to do that. I'm due to build a 4.12 package with the newer sa-exim soon. > Is there an easy way for me to rebuild the package to just depend on > the stable libc6 - or is this something you could fix, Marc? It would > make the package more useful to those of us trying to stay with the > stable code. Right. If you go to: http://marc.merlins.org/linux/exim/exim4-debian.html look at "you can get the debian subtree here (just untar it in your exim source distribution dir, and type debian/rules binary)" > "testing" since I've always just run the "stable" distribution. If I > change to "testing" in "source.list", dselect will want to upgrade all > packages to the "testing" version, right? Ugh. I don't use dselect, I just use apt, but I'm pretty sure that as long as you don't do an dist-upgrade, you can point to testing, and only upgrade the pacakges you want (and their dependencies) That's what I do (although it's not technically supported, but it's worked fine for me the last 4 years I've been doing it) Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems & security .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ | Finger marc_f@merlins.org for PGP key From thomask at mtnns.net Thu Jan 9 10:03:15 2003 From: thomask at mtnns.net (Thomas Kinghorn) Date: Thu, 9 Jan 2003 12:03:15 +0200 Subject: [SA-exim] why wont my whitelist work Message-ID: <4625C59C329BC447AFFB52E7F8BFF27504FF8F72@protea.int.citec.net> Good morning. I am running SA & EXIM. Here is my problem: Whitelist in local.cf does not seem to be working. Shouldn't the whitelisted address have 0 hits? I have restarted the spamd PID, no luck PID is: spamd -a -x -u xadmin Received: from mogwai.mtnns.net ([209.212.109.209]) by protea.int.citec.net with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2653.13) id YRR4NRRC; Mon, 6 Jan 2003 09:15:06 +0200 Received: from protea.int.citec.net ([209.212.109.146])..... MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) X-SA-Exim-Rcpt-To: thomas@loggies.za.net Subject: test Content-Type: text/plain; charset="iso-8859-1" X-Spam-Status: No, hits=4.2 required=4.5 tests=EXCHANGE_SERVER,SPAM_PHRASE_03_05 version=2.43 X-Spam-Level: **** X-SA-Exim-Scanned: Yes local.cf contains : whitelist_from_rcvd *@mtnns.net MTNNS.NET whitelist_from_rcvd *@citec.net CITEC.NET Regards, Tom Kinghorn From lists at timj.co.uk Thu Jan 9 12:04:17 2003 From: lists at timj.co.uk (Tim Jackson) Date: Thu, 9 Jan 2003 12:04:17 +0000 Subject: [SA-exim] why wont my whitelist work In-Reply-To: <4625C59C329BC447AFFB52E7F8BFF27504FF8F72@protea.int.citec.net> References: <4625C59C329BC447AFFB52E7F8BFF27504FF8F72@protea.int.citec.net> Message-ID: <20030109120417.5fa586e5.lists@timj.co.uk> Hi Thomas, on Thu, 9 Jan 2003 12:03:15 +0200 you wrote: > Whitelist in local.cf does not seem to be working. > local.cf contains : > whitelist_from_rcvd *@mtnns.net MTNNS.NET Try: whitelist_from *@mtnns.net That adds -100 points to the mail score. Tim From dman at dman.ddts.net Sat Jan 11 23:39:10 2003 From: dman at dman.ddts.net (Derrick 'dman' Hudson) Date: Sat, 11 Jan 2003 18:39:10 -0500 Subject: [SA-exim] Re: Debian issue w/exim4_i386.deb, depends on libc6 2.2.5-14.3 In-Reply-To: <20030107193814.GK4197@merlins.org> References: <20030107132708.B19086@doit.wisc.edu> <20030107193814.GK4197@merlins.org> Message-ID: <20030111233910.GA20973@dman.ddts.net> On Tue, Jan 07, 2003 at 11:38:14AM -0800, Marc MERLIN wrote: | On Tue, Jan 07, 2003 at 01:27:08PM -0600, Dave Plonka wrote: | > wants 2.2.5-14.3, which happens to be the current version from "testing" | > distribution. | | Ah, yeah, darn. | If someone who knows debian can tell me how to force an older version of the | libc when I build the package, I'll be happy to do that. The only way I know of is to have the desired version installed on the system. That can be done in a chroot jail, though. | > "testing" since I've always just run the "stable" distribution. If I | > change to "testing" in "source.list", dselect will want to upgrade all | > packages to the "testing" version, right? Ugh. | | I don't use dselect, I just use apt, but I'm pretty sure that as long as you | don't do an dist-upgrade, you can point to testing, and only upgrade the | pacakges you want (and their dependencies) | That's what I do (although it's not technically supported, but it's worked | fine for me the last 4 years I've been doing it) See apt_preferences(5). -D -- Even youths grow tired and weary, and young men stumble and fall; but those who hope in the Lord will renew their strength. They will soar on wings like eagles; they will run and not grow weary, they will walk and not be faint. Isaiah 40:31 http://dman.ddts.net/~dman/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 240 bytes Desc: not available Url : http://lists.merlins.org/archives/sa-exim/attachments/20030111/bcdd33a4/attachment.bin From marc at merlins.org Thu Jan 23 08:27:50 2003 From: marc at merlins.org (Marc MERLIN) Date: Thu, 23 Jan 2003 16:27:50 +0800 Subject: [SA-exim] Re: SA-Exim Problem In-Reply-To: <610F9B9BDB0FD41192BA006008551493039C0C@GANDALF> References: <610F9B9BDB0FD41192BA006008551493039C0C@GANDALF> Message-ID: <20030123082749.GF4633@merlins.org> On Wed, Jan 22, 2003 at 10:45:44PM -0000, Philip White wrote: > Hello Marc. > > Sorry to bother you, but I have a small problem I want to run past you. Try to use the mailing list, some people there can almost help you better than I can :-) (I'm at linux.conf.au in Australia, so I'll have to answer briefly) > Am runnimg Exim 4.12, with both Exiscan and SA-Exim compiled in. SA-Exim > seems to be running fine. Recognises and adds headers as expected. Ok. > Problem. > Exim acts as a MX for a hubbed host - all mail gets sent to an NT box > running MS Exchange, which delivers mail into a POP-box. I then collect mail > from this using KDEmail. > Somewhere along the line, the X-Spam headers are being removed, dropped, or > something. Well, it's very unlikely that exim has anything to do with it, especially if you've confirmed that sa-exim works correctly. exchange can remove headers it doesn't like AFAIK > The headers ARE there on the MS box. If I use Lookout on a Winbox, I can see > them. outlook talks to exchange via a special protocol IIRC, so exchange *could* give you a different view of the same mail via pop > Using SA through Exiscan works fine - the headers remain. But with SA-Exim, > I get everything up to and including the Subject:, then all else is lost. > > Question; > Are you aware of this happening anywhere else? > Any ideas on what is wrong? If you see the mails with the headers coming out of sa-exim when you read the mail locally (from spool, or from /var/spool/exim/SAspamaccept), the only thing that could happen in exim is headers_remove in your smtp transport or something. Marc -- "A mouse is a device used to point at the xterm you want to type in" - A.S.R. Microsoft is to operating systems & security .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ | Finger marc_f@merlins.org for PGP key From dman at dman.ddts.net Fri Jan 24 02:43:17 2003 From: dman at dman.ddts.net (Derrick 'dman' Hudson) Date: Thu, 23 Jan 2003 21:43:17 -0500 Subject: [SA-exim] Re: SA-Exim Problem In-Reply-To: <20030123082749.GF4633@merlins.org> References: <610F9B9BDB0FD41192BA006008551493039C0C@GANDALF> <20030123082749.GF4633@merlins.org> Message-ID: <20030124024317.GA1640@dman.ddts.net> On Thu, Jan 23, 2003 at 04:27:50PM +0800, Marc MERLIN wrote: | On Wed, Jan 22, 2003 at 10:45:44PM -0000, Philip White wrote: | > The headers ARE there on the MS box. If I use Lookout on a Winbox, | > I can see them. | | outlook talks to exchange via a special protocol IIRC, so exchange | *could* give you a different view of the same mail via pop Wouldn't that be nice =p? Ever heard the phrase "embrace and extend"? (not wholly applicable, but the general attitude is) If the headers are on the exchange system then the problem is almost surely not with (sa-)exim. | > Using SA through Exiscan works fine - the headers remain. But with SA-Exim, | > I get everything up to and including the Subject:, then all else is lost. | > | > Question; | > Are you aware of this happening anywhere else? | > Any ideas on what is wrong? | | If you see the mails with the headers coming out of sa-exim when you | read the mail locally (from spool, or from | /var/spool/exim/SAspamaccept), the only thing that could happen in exim | is headers_remove in your smtp transport or something. I suggest running some packet sniffer (eg tcpdump or ethereal) on the link to see what exactly passes from exim to exchange. If you see all the data passed on correctly then the problem is somewhere in exchange or beyond. Otherwise look deeper into (sa-)exim. HTH, -D -- In his heart a man plans his course, but the Lord determines his steps. Proverbs 16:9 http://dman.ddts.net/~dman/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 240 bytes Desc: not available Url : http://lists.merlins.org/archives/sa-exim/attachments/20030123/e5176d7a/attachment.bin From brian at enchanter.net Sun Jan 26 16:24:53 2003 From: brian at enchanter.net (Brian Kendig) Date: Sun, 26 Jan 2003 11:24:53 -0500 Subject: [SA-exim] FYI: Tarpit stats Message-ID: I was curious as to how effective my tarpit was being, so I grabbed my mail server log files and did some analysis on 'em. 1160 spammers escaped my tarpit (either by bailing out or waiting for the full time) during the past ten full days (January 15 through 25). Here's how long they spent in 'penalty hold': 950 were still stuck after half a minute 934 after a minute 918 after two minutes 864 after five minutes 848 after ten minutes 686 after twenty minutes 670 after half an hour 604 after an hour 559 after two hours 510 after four hours 494 at half a minute before five hours 348 at half a minute after five hours 307 at eight hours 243 at sixteen hours 192 at one day 129 at two days 91 at three days 66 at four days (345600 seconds, my limit, so I dumped the rest of 'em now) What this means is that roughly one out of every eighteen spammers is willing to sit around infinitely (four days is practically an infinity!) until I dump him. Note that about 150 of them disconnect right around five hours. Looks like some spammer software is using that as a timeout. At this moment I've got 39 spammers in my tarpit. I don't think I've ever had more than sixty or so simultaneously in the pit, even with the long delay and even though my server's configured to be able to handle 200 at once. I'm putting up these stats in case anyone's interested in them. :) ____ |\/| Brian Kendig Set your priorities right. \ /\ / ..__. brian at enchanter net No one ever said on his \/ \__\ _/ http://www.enchanter.net/ death bed, "Gee, if I'd \__ __ \_ Be insatiably curious. only spent more time at \____\___\ Ask "why" a lot. the office." From max at clarksys.com Tue Jan 28 18:42:28 2003 From: max at clarksys.com (Max Clark) Date: Tue, 28 Jan 2003 10:42:28 -0800 Subject: [SA-exim] Spamassassin: perl vs. spamc/d benchmarks Message-ID: Hi all! Does anyone know of any benchmarks between the command line perl and the spamc/d implementations of spamassassin? Do I lose any major features using spamc/d? Thanks in advance, Max From mboyd at cirilium.com Tue Jan 28 19:35:12 2003 From: mboyd at cirilium.com (Mark Boyd) Date: Tue, 28 Jan 2003 12:35:12 -0700 Subject: [SA-exim] Spamassassin: perl vs. spamc/d benchmarks Message-ID: <551810A11AE6DD4B81B4EE2C0E131CB0638CA1@typhoon.cirilium.com> I believe the web site shows some stats. If remember correctly for small e-mails it's a 5x improvement to use spamd/spamc but for larger ones only a 2x improvement. Check out the web site there's more information there: http://marc.merlins.org/linux/exim/sa.html - Mark -----Original Message----- From: Max Clark [mailto:max@clarksys.com]=20 Sent: Tuesday, January 28, 2003 11:42 AM To: sa-exim@lists.merlins.org Subject: [SA-exim] Spamassassin: perl vs. spamc/d benchmarks Hi all! Does anyone know of any benchmarks between the command line perl and the spamc/d implementations of spamassassin? Do I lose any major features using spamc/d? Thanks in advance, Max _______________________________________________ SA-Exim mailing list SA-Exim@lists.merlins.org http://lists.merlins.org/lists/listinfo/sa-exim