IMAP support

2004-03-19 Thread Daniel Perry



Can anyone elaborate 
on James's IMAP support?
The website doesnt 
say much.

I know it's not 
going to be classed as stable, but is it usable?


Is anyone 
sucessfully using IMAP?

I'm using james as 
our mail server, and people want to be able to use IMPs folders. IMP works 
fine with POP3, but doesnt have sent items/deleted 
items/etc.

Thanks,

Daniel.



Abuse or Configuration

2004-03-19 Thread Brandon Smith










I am having major problems with James.
Every minute a new log file is generated. I only found this out last night when
my server notified me that I was almost out of disk space. I noticed in about
three hours nearly 30 GB of log files had been generated. I shut down James in
an effort to investigate.



The two log files that are the culprits
are the spool and mailet log files. A mailet is generating errors and logging
each ToRepository. See below Listing One five entries of one of the mailet log
files. Youll notice that the timestampt is all 10:25:34. There is actually
about 150 entries each second. It takes several minutes (15-20) for a mailet
log file to get to max capacity and start a new log file. The spool log file is
much different.



The spool log file is generating the error
found in Listing Two nearly 150 times a second. Once a minute a new 10 MB file
is created. In the course of writing this email 6 where generated. What could
be the problem? Is James in a loop? Is my James being abused by a spammer? I
cant figure it outPlease help!





Listing One:



19/03/04 10:25:34 INFO James.Mailet:
ToRepository: Storing mail Mail1079693860218-98 in file://var/mail/error/

19/03/04 10:25:34 INFO James.Mailet:
ToRepository: Storing mail Mail1079693983078-99 in file://var/mail/error/

19/03/04 10:25:34 INFO James.Mailet:
ToRepository: Storing mail Mail1079693860218-98 in file://var/mail/error/

19/03/04 10:25:34 INFO James.Mailet:
ToRepository: Storing mail Mail1079693983078-99 in file://var/mail/error/

19/03/04 10:25:34 INFO James.Mailet:
ToRepository: Storing mail Mail1079693983078-99 in file://var/mail/error/





Listing Two:



19/03/04 10:24:54 ERROR spoolmanager:
Exception in JamesSpoolManager.run sun.io.ByteToCharEUC_CN.getIndex1()[S

java.lang.NoSuchMethodError:
sun.io.ByteToCharEUC_CN.getIndex1()[S

 at
sun.nio.cs.ext.EUC_CN$Decoder.init(EUC_CN.java:50)

 at
sun.nio.cs.ext.EUC_CN.newDecoder(EUC_CN.java:38)

 at
java.lang.StringCoding$CharsetSD.init(StringCoding.java:166)

 at
java.lang.StringCoding$CharsetSD.init(StringCoding.java:157)

 at
java.lang.StringCoding.decode(StringCoding.java:213)

 at
java.lang.String.init(String.java:320)

 at
javax.mail.internet.MimeUtility.decodeWord(MimeUtility.java:741)

 at
javax.mail.internet.MimeUtility.decodeText(MimeUtility.java:493)

 at
org.apache.james.core.MimeMessageWrapper.getSubject(MimeMessageWrapper.java:508)

 at
org.apache.james.transport.mailets.NotifyPostmaster.service(NotifyPostmaster.java:182)

 at
org.apache.james.transport.LinearProcessor.service(LinearProcessor.java:413)

 at
org.apache.james.transport.JamesSpoolManager.process(JamesSpoolManager.java:436)

 at
org.apache.james.transport.JamesSpoolManager.run(JamesSpoolManager.java:366)

 at
org.apache.avalon.excalibur.thread.impl.ExecutableRunnable.execute(ExecutableRunnable.java:47)

 at
org.apache.avalon.excalibur.thread.impl.WorkerThread.run(WorkerThread.java:80)










RE: Abuse or Configuration

2004-03-19 Thread Noel J. Bergman
Which JVM?  If 1.4.2, you *must* update to the very latest.  There was a bug
in all other versions of 1.4.2 that caused a decode error.

In any event, the loop is being caused by an exception in the error
processor.  If you use the current James test build, you can use this:

 mailet match=All
 onMailetException=ignore
 class=NotifyPostmaster
...
 /mailet

You won't get the notification, but you won't get the loop, either.
Actually, you should not get a loop like that, period.  I think that was a
bug we've already fixed.

--- Noel

-Original Message-
From: Brandon Smith [mailto:[EMAIL PROTECTED]
Sent: Friday, March 19, 2004 12:45
To: 'James Users List'
Subject: Abuse or Configuration


I am having major problems with James. Every minute a new log file is
generated. I only found this out last night when my server notified me that
I was almost out of disk space. I noticed in about three hours nearly 30 GB
of log files had been generated. I shut down James in an effort to
investigate.

The two log files that are the culprits are the spool and mailet log files.
A mailet is generating errors and logging each ToRepository. See below
Listing One five entries of one of the mailet log files. You'll notice that
the timestampt is all 10:25:34. There is actually about 150 entries each
second. It takes several minutes (15-20) for a mailet log file to get to max
capacity and start a new log file. The spool log file is much different.

The spool log file is generating the error found in Listing Two nearly 150
times a second. Once a minute a new 10 MB file is created. In the course of
writing this email 6 where generated. What could be the problem? Is James in
a loop? Is my James being abused by a spammer? I can't figure it out.Please
help!


Listing One:

19/03/04 10:25:34 INFO  James.Mailet: ToRepository: Storing mail
Mail1079693860218-98 in file://var/mail/error/
19/03/04 10:25:34 INFO  James.Mailet: ToRepository: Storing mail
Mail1079693983078-99 in file://var/mail/error/
19/03/04 10:25:34 INFO  James.Mailet: ToRepository: Storing mail
Mail1079693860218-98 in file://var/mail/error/
19/03/04 10:25:34 INFO  James.Mailet: ToRepository: Storing mail
Mail1079693983078-99 in file://var/mail/error/
19/03/04 10:25:34 INFO  James.Mailet: ToRepository: Storing mail
Mail1079693983078-99 in file://var/mail/error/


Listing Two:

19/03/04 10:24:54 ERROR spoolmanager: Exception in JamesSpoolManager.run
sun.io.ByteToCharEUC_CN.getIndex1()[S
java.lang.NoSuchMethodError: sun.io.ByteToCharEUC_CN.getIndex1()[S
at sun.nio.cs.ext.EUC_CN$Decoder.init(EUC_CN.java:50)
at sun.nio.cs.ext.EUC_CN.newDecoder(EUC_CN.java:38)
at
java.lang.StringCoding$CharsetSD.init(StringCoding.java:166)
at
java.lang.StringCoding$CharsetSD.init(StringCoding.java:157)
at java.lang.StringCoding.decode(StringCoding.java:213)
at java.lang.String.init(String.java:320)
at
javax.mail.internet.MimeUtility.decodeWord(MimeUtility.java:741)
at
javax.mail.internet.MimeUtility.decodeText(MimeUtility.java:493)
at
org.apache.james.core.MimeMessageWrapper.getSubject(MimeMessageWrapper.java:
508)
at
org.apache.james.transport.mailets.NotifyPostmaster.service(NotifyPostmaster
.java:182)
at
org.apache.james.transport.LinearProcessor.service(LinearProcessor.java:413)
at
org.apache.james.transport.JamesSpoolManager.process(JamesSpoolManager.java:
436)
at
org.apache.james.transport.JamesSpoolManager.run(JamesSpoolManager.java:366)
at
org.apache.avalon.excalibur.thread.impl.ExecutableRunnable.execute(Executabl
eRunnable.java:47)
at
org.apache.avalon.excalibur.thread.impl.WorkerThread.run(WorkerThread.java:8
0)


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Tracking each piece of mail through the system

2004-03-19 Thread Jerry Malcolm








Ive got a situation where some
legit mail is being categorized as spam. I run a few of the default spam
blockers that come with James. But I cant figure out which one is
causing a spam hit. Im familiar with all the various logs.
But there doesnt appear to be a single routing log that tracks a note
through all the processors/matchers/mailets.

Am I missing something? Is there an easy
way to enable some sort of tracking log?

Thanks.

Jerry










Planned v2.2.0a16 Test Build

2004-03-19 Thread Noel J. Bergman
The primary effort has been to merge the two main branches of James, but
there has been considerable other development on James.  The following are
expected for a test build that should be posted shortly.  There are more
changes already pending the merger.

--- Noel

 - Applied Apache License v2.0.
 - AbstractQuotaMatcher now handles  Sender properly
 - AbstractStorageQuota matcher now handles aliases properly
 - More information are made available through JMX.
 - Bugfix: http://nagoya.apache.org/jira/secure/ViewIssue.jspa?key=JAMES-152
 - Bugfix: http://nagoya.apache.org/jira/secure/ViewIssue.jspa?key=JAMES-151
 - Bugfix: http://nagoya.apache.org/jira/secure/ViewIssue.jspa?key=JAMES-150
 - Bugfix: http://nagoya.apache.org/jira/secure/ViewIssue.jspa?key=JAMES-149
 - MBoxMailRepository
   - list() verifies the first message before returning, just in case file
changed.
   - Changed interface to the callback object to explicitly indicate
 whether the operation is finished or not.
   - Changed to read lines instead of characters.  Both I/O
 mechanisms are still supported.  Jason Webb is working on a
 BufferedRandomAccessFile class, which preliminary tests suggest
 results in 20x performance improvement.
   - Added more DEEP_DEBUGGING messages to help with development.
   - Changed to rely upon the index.  findMessage will reload the
 keys (once) if selectMessage fails.
   - Pre-size the index based upon file size.
   - Use finally{} to ensure that files are closed.
   - Minor formatting fixes to log messages.
   - The mbox repository now uses two digit numeric days in the mbox
 mail separator. This will make the mbox repository more standards
 compliant and fix problems with some 3rd party IMAP/POP3 servers
 - FromRepository Mailet
   - Correct handling of boolean parameter
   - Add FromRepository mail attribute to make it easy to detect
 messages that have been spool from a repository.  THIS IS AN
 EXPERIMENTAL CHANGE.  Both the attribute name and the value
 should probably change.  Possibly the latter should be a URI
 for the repository or the message.
 - Bugfix: http://nagoya.apache.org/jira/secure/ViewIssue.jspa?key=JAMES-142
 - Added support for a flexible retry schedule to RemoteDelivery.
   This commit also includes a concurrent change to SpoolRepository so
   that accept() is a one-step process that returns the Mail, rather
   than an unsynchronized two-step process.  Another related change is
   that accept() takes a filter object to control the next message to
   be returned, rather than hardcoding the algorithm.
 - Fixed unexercised bug in File Repository code.


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]