[Bug 350341] Review Request: postfix-logwatch - A postfix log analyzer for logwatch

Tue Jan 29 17:58:30 UTC 2008

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug report.

Summary: Review Request: postfix-logwatch - A postfix log analyzer for logwatch
Alias: postfix-logwatch

https://bugzilla.redhat.com/show_bug.cgi?id=350341

------- Additional Comments From mike-bugzilla at cappella.us  2008-01-29 12:58 EST -------
I'm ok with whatever choice is made here.  But consider that 1.36.12 is fairly
old, and these changes have been made since:

2008-01-14 (version: 1.36.13pre7)
 - New: Support ETRN rejects (option: RejectEtrn).
 - Fix: Include optional text in L1 output for header/body checks Hold
   messages.
 - Fix: Accept "unknown" as an IP in Connection rate limit messages.
   Thanks: Stefan Jakobs.
 - Fix: Improve MxError captures
 - Fix: Ignore "sql auxprop plugin ..." messages
 - Fix: Ignore more debug_peer_level=2 messages
 - Fix: Correct failure to ignore "connect to subsystem..." debug lines
 - Fix: Add missing RejectVerify to config file

2007-12-15 (version: 1.36.13pre6)
 - Fix: in BounceLocal and EnvelopeSenderDomains, set null domain
   and formatted host to '*unknown'.
 - Fix: Move postfix_warning before postfix_cleanup, as some cleanup
   warnings were caught as unmatched.
 - Fix: An Accepted message is now triggered by smtpd "client=..."
   and pickup "uid=..." log entries, instead of qmgr's "from=xxx,
   size=nnn, nrcpt=nnn" log lines.  A message may have been accepted
   during a previous time period, but delivery delays result in
   multiple qmgr delivery attempts, resulting in over counting
   Accepted messages.  Thanks: Stefan Jakobs
 - New: Bytes delivered is now broken down by Bytes sent via SMTP,
   LMTP, and forwarded, to be orthogonal with message counts.
 - Fix: Resent messages no longer reduce the number of messages
   accepted - this was a naive attempt at not counting, for example,
   messages released and requeued from a content filter's quarantine.

2007-11-14 (version: 1.36.13pre5)
 - New: Threshold limiting and Top N lists for every level in each
   Detail section.  Every level in each section can now be limited
   with minimum count thresholds and top N lists.  See the updated
   README file, the comments in the postfix-logwatch.conf file, and
   the new postfix-logwatch man page.  Requested by: Pavel Urban
 - New: Rejects can now be categorized by reject reply code.  A new
   option/variable "reject_reply_patterns" is a list of reject reply
   code regular expressions, which are used for categorizing rejects.
   This feature allows, for example, distinguishing 421 transmission
   channel closes from 45x errors. (eg. 450 mailbox unavailable, 451
   local processing errors, 452 insufficient storage).  The default
   list is: "5.. 4.. Warn" which creates three groups of rejects:
   permanent rejects, temporary failures, and reject warnings (as in
   warn_if_reject).  Requested by: Noel Jones
 - New: postfix-logwatch man page created (net yet complete)
 - New: Support for all access(5) actions.  See the "Level Limiter
   Options" section in the postfix-logwatch(1) man page, and "Common
   access control actions" in postfix-logwatch.conf.
 - New: Added envelope senders and envelope sender domains reports.
   These are disabled by default.  Enable with level limiter options
   --envelopesenders 1 and --envelopesenderdomains 1 (or 2 to
   also see senders listed under domains).  See also the
   postfix-logwatch.conf file.  Suggested by: Brendan
 - Change: Uncomment all variables in the config file.  This should
   help ensure the variables in the config file stay in sync with
   those used in the source.
 - Change: Merged sections SenderDelayNotification, DSNDelivered, and
   DSNUndelivered into NotificationSent, with sub-sections indicating
   the type of notification.  This gives the total number of sent
   notifications in the Summary section, and the breakdown by type
   of notification in the Detailed section.
 - Change: Removed "msgs" prefix from several options: msgsdeferred,
   msgsdelivered, msgsforwarded, msgsresent, msgssent, and msgssentlmtp
   are now deferred, delivered, forwarded, resent, sent and sentlmtp.
   The old options are still usable.
 - Change: force --help output to 80 chars.
 - Change: Taint mode is now on by default in standalone mode.  It is
   disabled upon installation in logwatch mode, as logwatch fails with
   taint mode enabled.
 - Change: Set the primary key in tlsserverconnect/tlsclientconnect
   options to type/cypher to reduce excessive level 2 output.
 - Change: Reported values that cannot be determined or are unavailable
   are prefixed with an asterisk (e.g. *unknown, *unspecified).
 - Change: Warn section title 'Warn action logged' changed simply to
   "Warned", to be consistent with other access/header_ and body_checks.
   The option is --warned, but --warn is still acceptable.
 - Fix: Significantly reduce memory footprint when detail < 5.
 - Fix: Usage and --help now correctly show only detail section
   level limiter options that are available.  Previously, summary-
   only counts were also display as level limiter options.
 - Fix: "too many errors after DATA" and "timeout after DATA" may 
   include a byte count, as in "(348 bytes)".  Thanks: John Beaver
 - Fix: Ignore postgrey 'delayed ...' lines
 - Fix: Allow logwatch --debug option to pass into postfix-logwatch
 - Fix: configuration file reading code was not properly warning on
   non-existent files
 - Fix: "filter" actions were incremented on "redirect" actions
 - Fix: Give more room to percentiles in delays percentiles table.
   A 5 day delay is 432000.000 seconds, and the previous table width
   was not sufficient.
 - Fix: --show_sect_vars command line option inadvertently required an
   argument; the name has been shortened to --sect_vars/--nosect_vars
   and correctly no longer requires an argument.  The longer names
   --[no]show_sect_vars work as well.  Thanks: Noel Jones
 - Fix: Handle some unmatched connect to failures in section
   ConnectToFailure.  Cleanup and consolidate several similar messages;
   add additional detail at level 3.
 - Fix: Increment postsuper Hold messages by the number of messages
   placed on hold.
 - Fix: Add "non-ESMTP response ..." messages to SmtpConversationError.
   First level is now the general SMTP error description.
 - Internal: Converted source to be package-based to allow code sharing
   with amavis-logwatch.  The single-file executable is auto-generated
   from the packages.
 - Internal: internal gen_test_log function to create sample log data
   for testing. Reads '#TD' comments from within postfix-logwatch.

2007-10-16 (version: 1.36.13pre4)
 - Fix: Handle messages "SSL_connect error to example.com: 0" and
   "Cannot start TLS: handshake failure" by coercing into warnings
   Thanks: Rob Sterenborg

2007-10-15 (version: 1.36.13pre3)
 - Fix: Handle "postmaster" DSNs
 - Fix: Handle "discarding EHLO keywords:" (ignored)
 - Internal: create bounce subroutine to handle bounce messages

2007-10-05 (version: 1.36.13pre2)
 - New: Initial support for postgrey (http://postgrey.schweikert.ch/)
   Requested by: Sebastian Wolfgarten

2007-10-05 (version: 1.36.13pre1)
 - New: Support for policy-spf software postfix-policyd-spf-perl
   Requested by: Nicodemo P. and Rob Sterenborg

I can bump the version to 1.36.13 final if necessary - its been operational for
quite some time.

-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.