On Thursday 22 September 2005 12:18 pm, Justin Mason wrote: > Bowie Bailey writes: > > From: Dimitri Yioulos [mailto:[EMAIL PROTECTED] > > > > > On Thursday 22 September 2005 9:17 am, Bowie Bailey wrote: > > > > From: Dimitri Yioulos [mailto:[EMAIL PROTECTED] > > > > > > > > > As I understand it. DCC is no longer free, under some > > > > > circumstances, and is therefore disabled by default in 3.1.0. I > > > > > guess you have to manually enable it (although I don't know how > > > > > - can someone tell us how, please?). > > > > > > > > Take a look at the new v310.pre file. Uncomment the "loadplugin" > > > > lines for the ones you want to use. > > > > > > Thanks for tip! I guess I shoulda done more reading. That looks > > > like it should take care of the DCC and Razor problems. But what > > > about lock_method flock, and bayes breakage? I don't see anything > > > in that pre file realting to them. Unless ... using that pre file > > > instead of init.pre makes 3.1.0 happy and solves those problems. > > > Your thoughts? > > > > The way I understand it, the v310.pre is designed to work WITH > > init.pre. This way, whenever they add new plugins, they can just drop > > them in a new .pre file without having to worry about what you have > > changed in the old ones (i.e. When they release v3.2.0, it may add > > a v320.pre file). So, leave both .pre files there. > > Correct! > > > You could also migrate the extra v310.pre stuff into init.pre. Anyone > > else know if that would cause a problem down the line?. > > No, it will not, you could indeed do that (if you want). > > > As for the other problems... > > > > "lock_method flock" works fine for me on 3.1.0. I don't know why it > > wouldn't work for you. > > > > I don't have your original email anymore, so I don't remember what > > the bayes problem was. The machine I am using to test SA 3.1.0 does > > not have a trained Bayes DB, so I probably wouldn't be able to comment > > on that issue anyway. > > I suspect a typo -- it might be worth posting (a) your config file > and (b) output from "spamassassin --lint". > > --j.
Justin, Here's the config file (spam.assassin.prefs.conf): # MailScanner # MailScanner users, please see the comments at the bottom of this file. # MailScanner # # SpamAssassin user preferences file. # # Format: # # required_hits n # (how many hits are required to tag a mail as spam.) # # score SYMBOLIC_TEST_NAME n # (if this is omitted, 1 is used as a default score. # Set the score to 0 to ignore the test.) # # # starts a comment, whitespace is not significant. # ########################################################################### # JKF 12/01/2005 - known troublesome rule score ALL_TRUSTED 0 # JKF 21/12/2004 - relays.visi.com is dead score RCVD_IN_RSL 0 # JKF 1/12/2004 # If you use the phishing fraud detector, then setting this rule to a nice # high score will trap mail containing links to numeric IP addresses # (e.g. http://1.2.3.4/ instead of the normal http://www.domain.com). # score NORMAL_HTTP_TO_IP 4 # JKF 2/11/2004 # Added this setting so that SPF checks work properly with a default setup # of MailScanner. envelope_sender_header X-MailScanner-From ########################################################################### # First of all, the generally useful stuff; thresholds and the whitelist # of addresses which, for some reason or another, often trigger false # positives. # # JKF 28/04/2003 # The following settings has been pretty much superceded by the "Advanced # SpamAssassin Settings" in MailScanner.conf. # # JKF 26/03/2003 # If your root filesystem is filling up because SpamAssassin is putting # large databases in /.spamassassin or /root/.spamassassin, you can move # them using the following lines to point to their new locations. # The last part of the path is not a directory name, but actually the # start of the filenames. So with the settings below, the Bayes files will # be created as /var/spool/spamassassin/bayes_msgcount, etc. # auto_whitelist_path /etc/MailScanner/bayes/auto-whitelist auto_whitelist_file_mode 0600 bayes_path /etc/MailScanner/bayes/bayes bayes_file_mode 0777 # MailScanner: When using the scheduled Bayes expiry feature, you probably # MailScanner: want to turn off auto-expiry as it will rarely complete before # MailScanner: it is killed for taking too long. You will just end up with # MailScanner: big bayes_toks.new files wasting space. # bayes_auto_expire 0 # Whitelist and blacklist addresses are *not* patterns; they're just normal # strings. one exception is that "[EMAIL PROTECTED]" is allowed. They should be in # lower-case. You can either add multiple addrs on one line, # whitespace-separated, or you can use multiple lines. # # Monty Solomon: he posts from an ISP that has often been the source of spam # (no fault of his own ;), and sometimes uses Bcc: when mailing. # #whitelist_from [EMAIL PROTECTED] # Add your blacklist entries in the same format... # # blacklist_from [EMAIL PROTECTED] # Mail using languages used in these country codes will not be marked # as being possibly spam in a foreign language. # # By default, SpamAssassin will run RBL checks. If your ISP already # does this, set this to 1. # # skip_rbl_checks 1 ########################################################################### # Add your own customised scores for some tests below. The default scores are # read from the installed "spamassassin.cf" file, but you can override them # here. To see the list of tests and their default scores, go to # http://spamassassin.taint.org/tests.html . # MailScanner: Comment out the next line to enable DCC checking if you # have dcc installed (optional part of SpamAssassin) # JKF Commented out as it no longer generates maillog warnings #score DCC_CHECK 0.0 #dcc_path /usr/bin/dccproc #dcc_home /etc/dcc # # Added for MailScanner 23/5/2003 # The timeouts for blacklists and Razor are rather generous in the default # state that SpamAssassin is shipped. Reducing these stops a lot of timeouts # from removing SpamAssassin scores altogether. # rbl_timeout 20 razor_timeout 10 pyzor_timeout 10 # # Added for MailScanner 14/6/2002 # If you specify these scores, SpamAssassin will do RBL checks as well as # MailScanner, which just wastes CPU power and network bandwidth. Either # do them here by uncommenting the rules below (if you have paid for them) # or else uncomment the "skip_rbl_checks" line above and let MailScanner # do the checks instead. # #score RCVD_IN_BL_SPAMCOP_NET 4 # These next 3 will cost you money, see mailscanner.conf. #score RCVD_IN_RBL 10 #score RCVD_IN_RSS 1 #score RCVD_IN_DUL 1 # For spam and notspam bins bayes_ignore_header X-MailScanner bayes_ignore_header X-MailScanner-SpamCheck bayes_ignore_header X-MailScanner-SpamScore bayes_ignore_header X-MailScanner-Information # By default, the Bayesian engine is used. This is a real CPU hog and uses # a lot of system resources to work. # On a small overloaded system, you might need to disable it. use_bayes 1 # Most people don't use NFS-shared Bayes databases # so this is added for SpamAssassin 3 lock_method flock Output from spammassasin --lint (don't have with -D switch): Failed to parse line in SpamAssassin configuration, skipping: envelope_sender_header X-MailScanner-From Failed to parse line in SpamAssassin configuration, skipping: dcc_home /etc/dcc Failed to parse line in SpamAssassin configuration, skipping: lock_method flock Dimitri
