Spamtrainer and opendirectory

I have set up a mail server with mac os x 10.4 server (server2). The mailboxes are created on another mac os 10.4 server (server1) in an opendirectory enviroment. I have setup the junkmail and notjunkmail mailboxes on the opendirectory server. When I launch "sudo spamtrainer" the cpu goes 100% and in hours nothing happens. When I kill the process the spamtrainer.log shows something like this:
Sep 07 11:10:05 - spamtrainer Version 1.7.2
0.000 0 18897 0 non-token data: ham
0.000 0 3685 0 non-token data: spam
this is how i setup spamtrainer
Checking if there is a startup item for 'learnjunkmail' or 'spamtrainer'
There IS NO plist for learnjunkmail
There IS NO cron job for learnjunkmail
There IS NO plist for spamtrainer
There IS NO cron job for spamtrainer
if you want to use 'spamtrainer' it is recommended that this be added
Would you like me to enable it for you (yes/no)
yes
What time would you like the job to run (24-hour format)?
Please enter the hour (1)
23
Please enter the minutes (0)
00
Would you like SPAM/HAM messages to be deleted after the learn process? (n)
y
Would you like to log bayes stats into /var/log/spamtrainer.log? (n)
y
Enter name of mail store with SPAM/HAM mailboxes? Enter for default (default)
Enter name of mailbox with SPAM? Enter for default (junkmail)
Enter name of mailbox with HAM? Enter for default (notjunkmail)
The following cron job for 'spamtrainer' has been prepared
00 23 * * * /usr/sbin/spamtrainer -d -l
Would you like to add it? (yes/no)
yes
The following cron job for 'spamtrainer' has been enabled'
00 23 * * * /usr/sbin/spamtrainer -d -l
anyone can help???

Well to start off with, grab a copy of Siradmin and using the instructions included in the help file make an admin user of your own choosing (i use mailwitch) and add her to the cyrus admin list (see the help file). Then open up the mailboxes on the second server and see if they (junk and not junk) are there. If they are - open up permissions on them so can get mail and are addressable by anyone (sometimes my IMAP users drag and drop mail to the respective boxes instead of doing re-directs)- Admin's choice.
As to the OD question, you;ll know if the mailboxes are local when you open up siradmin. I believe thats something you'd have to spec in OD.YMMV. You should also try Alex(the author of spamtrainer) he's fantastic, patient speedy to respond and frequently seen here.
My $.02
Sean

Similar Messages

  • 10.6.8 server, spamtrainer and sa-learn?

    When I run spam trainer I get:
    Starting spamtrainer...
    Training from user folders
    Learning SPAM...
    Learned new SPAM (junk mail)
    Learning HAM...
    Learned new HAM (not junk mail)
    Syncing SpamAssassin Database
    Displaying SpamAssassin Database Stats
    0.000          0       1564          0  non-token data: spam
    0.000          0       8824          0  non-token data: ham
    Deleting learned HAM and SPAM
    Done!
    Output produced by spamtrainer Version 1.9.7
    But then if I run sa-learn --dump magic right after it, I see different totals.
    bash-3.2# sa-learn -u _amavisd --dump magic
    0.000          0          3          0  non-token data: bayes db version
    0.000          0        354          0  non-token data: nspam
    0.000          0       1491          0  non-token data: nham
    0.000          0      90650          0  non-token data: ntokens
    0.000          0 1029372712          0  non-token data: oldest atime
    0.000          0 1314672396          0  non-token data: newest atime
    0.000          0 1314672741          0  non-token data: last journal sync atime
    0.000          0          0          0  non-token data: last expiry atime
    0.000          0          0          0  non-token data: last expire atime delta
    0.000          0          0          0  non-token data: last expire reduction count
    Can someone explain what's going on?

    I've seen this error generally when Server Admin is left open for long periods of time on the server itself. It tends to become unstable for no apparent reason. Rarely see it when using the application remotely in my experience. I've also noticed this error when using the server's bonjour name (which I tend to avoid) rather than its fully qualified domain name (which is what I tend to do). Server Admin should never run like a one legged cat and if it does I would double-check DNS first, followed by permissions and/or the OS and/or the system drive's 'health' afterwards. But without being there it would be difficult to diagnose.
    Guidelines on how to set-up an OD Master/Replica relationship should be outlined in the Snow Leopard Server Manual:
    https://manuals.info.apple.com/MANUALS/1000/MA1180/en_US/OpenDirAdmin_v10.6.pdf
    Chapter 4. Failing that I would look at the Open Directory Admin Manual for Tiger Server (10.4) because as I recall the subject is covered in more detail? It's been a while but basically both servers must be exactly the same OS, DNS as it should be, clocks are synced and SSH is enabled. Beyond that root should be used. Should be the same process in Yosemite although I would check with the Advanced Server Manual for 10.10 first:
    https://help.apple.com/advancedserveradmin/mac/4.0/#/apdDC3BA60F-246B-43E8-AB24- F6419B32FF6D
    It appears to be fairly limited in in the 'how' of OD Master/Replica relationships? Looking at the Server.App for Yosemite I'm thinking it's all done there?

  • Spamtrainer and future dated spam

    I'm running:
    Mac OS X Server 10.4.11
    Updated the components from osx.topicdesk.com
    Cron job for spamtrainer
    I noticed that spam with a date in the future does not get processed and erased when spamtrainer is run by the cron job.
    Is this normal?
    Should I just delete the future dated spam from my junk mailbox?
    Thanks

    It is normal and those messages will be used for training. Just not deleted. You can either leave them for future deletion or delete them manually. Doesn't really matter.

  • Spamtrainer & sa-learn

    I was just going through a recent test server (10.4.7), which had previously not had spamassassin enabled, in order to get a clearer idea of spamassassin and its learning process.
    I fixed the .spamassassin link, added the 2 junkmail users, installed spamtrainer. Ran spamtrainer and it reported (correctly) just the couple of messages I had used to test the spam & ham accounts.
    I ran sa-learn on two folders, one spam and one ham, of 400 messages each.
    I then redirected an obvious spam message to a user account. This failed to be labeled as spam, with the header stating "X-Spam-Status: No. Hits: 0. Tests: <blank>"
    I redirected the message to the junkmail account and ran spamtrainer which reported that the number of messages in the Spam database was now 3. Not the 403 which I kind of expected after the sa-learn process (I obviously have the wrong idea about what is being reported in the table).
    Amavis.log shows the scan going through its motions & running a manual test on the text of the message... " spamassassin -D < spam.txt " displays a test score of 0.2 so at least it seems to be being scanned.
    But why has the sa-learn process, on 400 spam & ham, not seemed to have had any effect? I can only presume I have missed something obvious...
    Ta.
    -david
    Size of files in .spamassassin folder...
     172032 auto-whitelist
      90112 bayes_seen
    1331200 bayes_toks
       1175 user_prefs

    Thanks, Alex.
    I carried out a further couple of tests on a purged database (Server V. 10.4.7)...
    Test 1. Purge database then test 4 pieces of spam.
    Test 2. Purge, sa-learn on 500 each of spam and ham, test same 4 spam.
    The following rules & scores were reported for each test...
    SPAM1
    TEST1: 0.2 no real name, 0.2 invalid date, 0.1 forged rcvd helo. TOTAL 0.5
    TEST2: 0.0 no real name, 0.2 invalid date, 0.4 bayes 60. TOTAL 0.6
    SPAM2
    TEST1: 0.2 no real name, 0.2 invalid date, 0.1 forged rcvd helo. TOTAL 0.5
    TEST2: 0.0 no real name, 0.2 invalid date, 0.0 bayes 50. TOTAL 0.2
    SPAM3
    TEST1: 0.0 missing date, -2.8 all trusted, 2.0 head illegal chars, 1.6 missing subject. TOTAL 0.8
    TEST2: -3.3 all trusted, 2.1 head illegal chars, 0.1 missing headers, 1.2 missing subject. TOTAL 0.2
    SPAM4
    TEST1: 0.2 no real name, 0.2 invalid date, 0.5 info TLD. TOTAL 0.9
    TEST2: 0.0 no real name, 0.2 invalid date, 1.9 bayes 99, 0.0 mime qp long line. TOTAL 2.2
    Conclusions?
    There is obviously a lot of dynamic adjustments to the scores given to non-body rules which are triggered by passing the bayes threshold of 200 messages. I also tested after training with spam only and there was no changes shown to the test scores compared to the tests run on a purged database.
    Although I should have scanned also a sample of ham, I'm beginning to think that initial bulk training, particularly from a '3rd party' source, may be of insignificant benefit at best and actually counter-productive at worst. Although the initial training triggers bayes analysis, the subject of newly incoming 'real' mail is likely to be so different from my own sample of spam/ham (particularly ham) that the bayes statistics will be commencing from a skewed basis. It would probably need just as many (if not more) real samples of ham in order to start accurately assessing the incoming text bodies.
    On the same basis, ongoing training from a third party spam source would also arguably be counter-productive.
    So it seems that there is not really any alternative to in-house training (which was the question I was first trying to answer when I started looking at this closer).
    As you have pointed out, Alex (and others in the forum), in order to commence from a more effective starting point then the addition of more rules - rather than more training data - seems the productive way to go. I'll implement URIBL/Sare rules mentioned in other threads and, just out of curiosity, report back on the same tests.
    Thanks again for feedback.
    -david

  • Has anyone integrated OpenID into the Mac OS X Server Wiki and Blog?

    I am currently working on integrating OpenID into the eXist Open Source XML Database (http://exist-db.org). As a separate topic, I was wondering if anyone has integrated OpenID into the Wiki, Blog and OpenDirectory.

    do
    you
    have
    any
    adviceon
    the
    best
    way
    to
    do
    this?
    or
    can
    you
    guide
    me
    to
    a
    beginners
    guide
    on
    how
    to
    set
    up
    a
    Web
    page?

  • VPN Causes Xserv to slowdown and stops services

    Hey guys,
    I have noticed when someone vpn's into the server it causes the other services that are running to slow down and gives the spinning domino's at the bottom of the server admin screen and all of the services that are indicated by the green buttons get hyphens on them. Everything on the server is movable, like mouse so I know the server is not locked up but I can notice that when someone vpn's it causes the services to go slower and causes the mail to give errors on people on our network. Any suggestions??
    G5 Xserv   Mac OS X (10.4.9)  

    VPN clients are given 10.20.100.201 - 209
    DHCP clients are given 10.20.100.10 - 200
    Servers are >.10 and printers are <.210 for our office.
    I set the routing definition to 10.20.100.0 // 255.255.255.0 // private
    The server has 3 NIC cards inside it.
    The Built in NIC (eth0) is serving the 10. subnet with an address of 10.20.100.1
    eth1 is an apple gig-e card, connected to our SDSL line with a static IP
    eth2 is an apple gig-e card, connected to nothing.
    I am able to connect SMB shares on the server when connected to the VPN by entering their implicit path in windows //server01/sharedocs, or by afp://10.20.100.1 once the connection is made.
    While I remain connected the server is still allowing mail & WebDav connections to our Kerio mail, and webmail users can still get mail. However within a few minutes of disconnecting no one is able to get mail either through entourage, mac mail, outlook or webmail. I am at that point unable to SSH or Timbuktu Pro connect to the server (they time out) i can bring up webmail, and log in, but it hangs at that point, I can bring up the Admin HTTPS page of the kerio server and view settings and status (0 clients connected)
    I am wondering if this isnt some issue between the VPN and OpenDirectory, because SSH uses the usernames in open directory, so is Timbuktu, the mail server is getting its users from open directory, so is WebDav, but the Admin account of the webmail is hardwired into the server and not pulled from OpenDirectory. I have nothing but that to go on, and i lack the knowledge to know where to look to test that theory.

  • URIBL and Bind

    I recently added URIBL checking to my Spamassassin as per this tutorial
    http://wiki.apache.org/spamassassin/SpamAssassinon_Mac_OS_XServer
    Seemed to work 'ok'. But then later in the week I needed to adjust a zone file for my domain (eg clienttestsite.mydomain.com). After restarting Bind and confirming that it was running and resolving fine, my mail service seemed to halt all mail delivery with messages such as:
    relay=none, delay=300, status=deferred (connect to 127.0.0.1[127.0.0.1]: read timeout)
    Mar 12 11:37:27 resserver postfix/smtp[2166]: connect to 127.0.0.1[127.0.0.1]: read timeout (port 10024)
    Note: I'm running mail and primary DNS services on this box.
    I suspect that mail wasn't being 'okay-ed' by amavisd because the URIBL checks were failing to resolve when the DNS was down. It appeared that amavis just locked up untill the mail service was restarted.
    I temporarily removed the content_filter line from my postfix conf and retried sending out of my queue which worked.
    I suspect that if I put amavisd back in my postfix conf that things will start working again, but I'm afraid of this happening again when I need to edit a zone (which is regularly).
    Is there a way I can list my backup DNS server in there somewhere so that if I'm restarting BIND on the Mail server it can use the backup DNS box for it's lookups until the DNS is back up on itself?

    Seems like it was corrupted because I ran spamtrainer and it complained about all kinds of things missing once it got to my 'ham'. (the only thing I had changed before the last successful run was putting 1000+ messages into my notjunkmail account which previously had only 75 or so) Tried running spamtrainer again and it stuck on reading my spam folder forever, I finally killed it after 3 days.
    So I suspect that the db got corrupted somehow?
    Because of all of the content filter issues I've been having I currently have the 'content_filter' line in my main.cf temporarily commented out. We've been using RBL's and greylisting on the frontline which seems to block most of our spam, I am however interested in getting my content filters 'up to snuff' to get the few that get through. Is spamtrainer still able to do it's job bulding the db without that linkage to postfix? I'm assuming I can let spamtrainer run without this postfix linkage and then uncomment the line later when my content_filter stuff is configured properly and ready?

  • Lost access to junkmail and notjunkmail??

    I set up these two imap email folders with WGM and I could access them with Mail. Then because I wasn't satisfied with the spam filtering
    I installed spamtrainer and mailfbr and ran both. These both seem to be working, but now I can't access via Mail or web2mail the two junk mail accounts. They show up in WGM and Service Admin - mail and Mailfbr displays them when it runs, but no receiving mail or access??
    Like I said they both ran fine b/4 running spamtrainer and mailfbr, but not afterwards. Anyone have suggestions to fix? Or what happened?

    By using mailbfr -f, you definitely did lose your custom ACLs.
    (I'm making an assumption below)
    Previously, you had ACLs setup which shared junkmail and notjunkmail to you and/or other users.
    The easiest way to restore the ACLs is to download SirAdmin:
    http://www.apple.com/downloads/macosx/networking_security/siradmin.html
    Use this tool to add an ACL to both junkmail and notjunkmail. You can specify how much access you should have (full read/write, write only, etc). I can't tell you what is right for you. Some clients provide ALL access to junkmail/notjunkmail, but that introduces some privacy issues. Some clients have notjunkmail (which contain legit mail) setup as a drop-box (ACLs = LIP) so that you can drop mail but not read what others have dropped.
    Start with SirAdmin and it's directions.
    Jeff

  • Using spamtrainer then Server Admin

    I've run spamtrainer and followed all the instructions included with it and all looks well...
    I believe I remember reading in some of those instuctions that I should stay clear of changing settings with Server Admin once I've using spamtrainer.
    Is it safe to change the "Minimum junk mail score" setting in Server Admin? What are the settings (or is it just any of them) that one should refrain from changing in Server Admin after running spamtrainer?
    By the way pterobyte, if you are the one that answers, thanks for all your posts and for spamtrainer.

    Alan,
    spamtrainer does NOT limit your use of Server Admin in any way whatsoever.
    What you probably read somewhere is that you should ditch Server Admin once you start manually modifying postfix and amavisd configuration files.
    Yes it is safe to change the score in SA.
    Alex
    P.S. You are welcome

  • How to set up Open Directory Users with local home folders?

    Hi folks,
    i set up a Mac mini Server with the services DNS, DHCP, AFP and OpenDirectory running. Everything is working fine so far but i want my OpenDirectory Users to have their home folders locally on the clients harddisk. My Leopard clients are already bound to the directory but everytime i try to login the login window is shaking even when i use the Directory Administrator account. What am i doing wrong?
    Thanks.

    (Did you read my other reply? You need to make it a portable account for the caching of login credentials.)
    With network homes and portable account enabled on a machine you always run from the locally stored homefolder on that machine.
    This homefolder syncs with the server network home folder for that account.
    If either of these folders get corrupt or unintentionally altered files (permission problems) somehow, there is risk of losing files, syncing problems and more (can take long time to sync at login/out, during even without corruption - I prefer gigabit cable to WiFi for this for obvious reasons).
    You at least need to monitor storage space in both places. You don't want either to fill up the disk (worse on server because several users can have their folders corrupted at the same time). Working quota settings / account is probably a good thing.
    And you probably need to decide what is synced : all folders or just some (and when / how often).
    You either do this form the server (you decide) or let the user have some say or a mix of these two.
    You might want to leave out the user's personal files (music, movies and such) and also depending on emaIl client used, maybe not sync email if the client saves it like Entourage does it, in big ever changing database files. Might be better to just backup the mailserver if IMAP based.
    If a user puts other large files in their homefolder (often on the Desktop) they can fill up the server fast. DVD-projects anyone?
    Mixing different versions server/client might not work too good either.
    So having said that, when syncing works it can be really good and you can restore a machine/account fast with pretty recent data. You can if in a hurry even log in from another machine using the account network home folder, getting access to the synced folders/files.
    Using Time Machine it's a one way "sync" (more like "duplicate changed and new files" - which I like) and you have to manually set TM settings on the client for what is going to be backed up and when. I do atleast don't know how to do it from/on the server. This "less intervention by the server" can be a good thing but if your users don't "behave" you might want to be in the drivers seat "saving them from themselves".
    TM backups can grow fast and probably demands more user interaction when storage space is used up.
    As with all "backup" configurations you probably want some rotating media backup of both network homes and/or TM backups so you can do a restore. Of these two, network homes are the more important one to backup "further".

  • Portable Home Directory no longer works on Lion bound to Active DirectoryAD

    Hi, All,
    I look after about 30 MacBook Air/Pro connected to a predominantly Windows corporate network.
    A few months back, I installed a MacOS Server running OpenDirectory and Software Update Services for managing the MacBooks - they are configured in the standard "magic triangle" configuration, bound both to the Active Directory (running on a handfull of Windows 2003-R2 servers) and OpenDirectory; they are configured to use the Mac Server for Software Updates, and with Portable Home Directories replicated on a network share point (through AFP) to keep a backup of the user's data.
    When all the MacBooks were running Snow Leopard, everything was running fine... but the last couple of laptops purchased came with Lion and they cannot synchronise their PHD's anymore :-(
    I have spent a lot of time trying to pin-point the problem, including replicating a whole test-network with its own AD/OD and client Snow-Leopard/Lion laptops to try various configurations... my conclusions so far are that:
    1- using Snow Leopard clients, it just works
    2- using a Lion client, when the network user account is configured in the default Users O.U. on Active Directory, it just works
    3- using a Lion client, when the network user account is configured in another O.U. than Users (e.g. Company-Name or External-Contractors), then the PHD syncinc doesn't work at all
    4- the failure symptoms are a popup window when trying to do a manual sync saying that "your network home at (null) does not allow writing" - if automatic syncing is configured through MCX, it just fails silently
    5- when PHD syncing fails, I can log on using the network account, manually mount the share point used for the network home in the Finder and read and write to it without a problem; so it' not a permission problem, but the fact that the network home directory is (null)
    6- I have traced the root cause to the FileSyncAgent process which raises an exception at startup - here's what I see as the salient lines from the logfile (i've edited out the username):
    0:: [12/02/07 12:45:00.512] ******************************************************************************
    0:: [12/02/07 12:45:00.512] FileSyncAgent-502.2 (r?, BUILT:?, PID:385, OS:11D50b, ARCH:x86_64-64) starting
    0:: [12/02/07 12:45:00.512] LA: FileSyncAgent -launchedByLaunchd -iDiskPlist
    0:: [12/02/07 12:45:00.512] ******************************************************************************
    0:: [12/02/07 12:45:00.512] Engineering log verbosity level = 1
    1:: [12/02/07 12:45:00.512] Registered isRunning port with name 'com.apple.FileSyncAgent.iDisk.isRunning'
    1:: [12/02/07 12:45:00.846] Temporary disk storage at "/Users/[USERNAME]/Library/Caches/Cleanup At Startup/FileSyncAgent-1727909307".
    1:: [12/02/07 12:45:00.846] UserAgentString = "DotMacKit-like, File-Sync-Direct/502.2.? (11D50b x86_64-64)"
    1:: [12/02/07 12:45:03.249] +[SSyncSet_PHD createPHDSyncSetForLocalPath:remoteHomeSpec:mountSuffixPath:name:]: Creating Sync Set with name "HomeSync_Mirror".
    1:: [12/02/07 12:45:03.249] -[SSyncSet_PHD _setupNewPHDSyncSetWithLocalPath:homeSpec:mountSuffixPath:]: We've been given local home path '/Users/[USERNAME]'.
    1:: [12/02/07 12:45:03.249] -[SSyncSet_PHD _setupNewPHDSyncSetWithLocalPath:homeSpec:mountSuffixPath:]: local home at "/Users/[USERNAME]"
    <PHD> 1:: [12/02/07 12:45:06.458] Added new sync set "HomeSync_Mirror".
    <PHD> 1:: [12/02/07 12:45:06.465] _incomingIPC: SFCreatePHDSyncSetMsgId (17) took 3.217627 seconds.
    <PHD> 1:: [12/02/07 12:45:06.526] Scheduling next sync of "HomeSync_Mirror" at 2012-02-07 12:45:16 +0000
    <PHD> 1:: [12/02/07 12:45:18.122] ==========================================================
    <PHD> 0:: [12/02/07 12:45:18.123] Starting automatic sync of "HomeSync_Mirror".
    <PHD> 1:: [12/02/07 12:45:18.126] Peer "local" reports changes since last sync.
    <PHD> 0:: [12/02/07 12:45:18.128] EXCEPTION: NilPtr <-[SPeer_FS_PHD mountPeerVolume] (Peer-FS-PHD.m:142): "'((homePath))' is nil">
    <PHD> 0:: [12/02/07 12:45:18.128] BACKTRACE: {
    <PHD> 0:: [12/02/07 12:45:18.128] ? | 0x105003493  
    <PHD> 0:: [12/02/07 12:45:18.128] ? | 0x104f70866  
    <PHD> 0:: [12/02/07 12:45:18.128] ? | 0x104f6fabd  
    <PHD> 0:: [12/02/07 12:45:18.128] ? | 0x104f6ecb4  
    <PHD> 0:: [12/02/07 12:45:18.128] ? | 0x7fff924bb74e
    <PHD> 0:: [12/02/07 12:45:18.128] ? | 0x7fff924bb6c6
    <PHD> 0:: [12/02/07 12:45:18.128] ? | 0x7fff901998bf
    <PHD> 0:: [12/02/07 12:45:18.128] ? | 0x7fff9019cb75
    <PHD> 0:: [12/02/07 12:45:18.128] }
    7- if I use the same user account from a Snow-Leopard client, the mountPeerVolume finds the remote home_dir, mounts it and happilly trawls through the two directories to figure out what needs to be copied; I get something along the lines of:
    0:: [12/02/08 18:44:15.344] Starting manual sync of "HomeSync_Mirror".
    1:: [12/02/08 18:44:15.363] -[SPeer_FS_PHD mountPeerVolume]: We've been given remote home path "/Volumes/[USERNAME]".
    1:: [12/02/08 18:44:15.363] -[SPeer_FS_PHD mountPeerVolume]: Remote home path exists.
    1:: [12/02/08 18:44:15.363] -[SPeer_FS_PHD mountPeerVolume]: Final path to PHD remote home root = "/Volumes[USERNAME]"
    8- if I use a Lion client and a user account configured in the Users O.U. on Active Directory, it works similarly well
    9- in the failing use-case, if I run a "dscl xxx -read /Users/xxx" to verify that the DirectoryServices deamon returns some information, all looks good: I get all my A.D. user details, including email config and network home_directory, the MCX payload, etc
    10- the only difference that I can see when running dscl between the "working use-case" and the "failing use-case" is that the Attribute for the network home directory is called HomeDirectory (working) or OriginalHomeDirectory (failing) - but the actual value is present and correct in both cases
    11- I have enabled SMB as well as AFP file sharing to ensure that it's not a protocol problem rather than the share point itself that causes problem - no difference
    12- I have upgraded the MacOS Server to 10.7.3 (from 10.7.2) - no difference
    13- I have upgraded the Lion client from 10.7.2 to 10.7.3 - no difference
    14- I have tried to manually configure the Directory Search path on the Lion client to map HomeDirectory to OriginalHomeDirectory - either the mapping didn't work or it's not the real cause of the problem, as again: no difference.
    15- I have contacted AppleCare, raised a case with all the details above, been escalated from level 1 to 2 and 3 - and basically have been told that customers have reported very similar problems & that no more detail is available to me as it is a Directory Services related problem - if I want to take it further with Apple, I need Enterprise OS Support (and hence $1000s per year - and no real guarantee of result)
    [thanks to anyone who's read until now]...
    So my questions are (as I can't believe that I'm the only one using "proper O.U.'s" for managing users in Active Directory):
    - has anyone experienced this before?
    - does someone know of a workaround (other than "stick to SnowLeo" or "redesign your whole A.D."!)?
    - can someone help me figure out where the FileSyncAgent process tries to get its remote home directory from - as this might point to a solution?
    Many thanks

    I was able to fix the issue here. Steps Taken:  From the comand line I ran
    dscl /Search read /Users/jdoe HomeDirectory
    It returned # No such key: HomeDirectory
    This was the big clue.... it should have returned
    HomeDirectory: <home_dir><url>afp://files.hihllc.lan/Users/jdoe</url><path>/</path></home_dir>
    I unbound the Mac from AD and ran the following command in terminal.
    dscacheutil -flushcache
    I then went into OS X sharing preferences and renamed the computer slightly, so it would create a new computer record in AD. I rebound the machine to AD making sure in Directory Utility Advanced, under User Experience that Force local home directory on start up disk was unchecked and Network protocol to be used was AFP.
    I then ran
    dscl /Search read /Users/jdoe HomeDirectory
    and it returned
    HomeDirectory: <home_dir><url>afp://files.hihllc.lan/Users/jdoe</url><path>/</path></home_dir>
    I had the user reboot... which was necessary. When he logged in HomeSync fired up. He had quite a few conflicts, so I told him to select use files on this computer for all conflicts.
    I checked the FileSyncAgent.logs and it was good.
    Let me know if this fixes your issue.
    Thanks,
    Ray

  • Printer does not show in Rendezvous in Preinter Setup Utility

    I am using a G4 Ti 15" laptop usually connected wirelessly to an AirPort Extreme BS. Also connected to that BS is a desktop (ethernet), cable modem, and a HP LaserJet 2200 printer. If I connected the printer directly to the laptop, it prints. However, if I try to print wirelessly, it doed not work. In the Printer Setup Utility when I try to add a printer and select Rendezvous, the printer does not showup in the available list.

    Erven,
    Have you followed these instructions from the help pages of Printer Setup Utility?
    Connecting to a network printer
    If your computer is connected to a network, you can use the printers in your local area, including those that use Bonjour (formerly Rendezvous), AppleTalk, and OpenDirectory.
    Open a document to print, and then choose File > Print.
    Choose Add Printer from the Printer pop-up menu, select the printer in the dialog that appears, and then click Add.
    The dialog lists shared, Bonjour, AppleTalk, and Open Directory printers. You may need to wait a short while before your printer appears.
    If you don't see your printer, click More Printers, choose your printer's type from the pop-up menu in the dialog that appears, select the printer, and then click Add.
    If the specific type of printer you're using is listed, choose that type. For example, if you're using an Epson AppleTalk printer, you may be able to choose Epson AppleTalk.
    If you choose AppleTalk, another pop-up menu appears that lets you choose the AppleTalk zone for the printer.
    When you select your printer in the Add Printer or More Printers dialog, Mac OS X finds the printer's driver and adds the printer to the Printer pop-up menu. If Mac OS X cannot the printer driver, "Driver not Installed" or "Gimp-print" appears in the model pop-up menu.
    If you still cannot find the printer, it may be an IP printer, and you will need to follow the instructions for how to set up an IP printer.
    Usually, your computer can detect whether a printer has special accessories installed, such as additional paper trays, extra memory, or a duplex unit. If it can't, a dialog appears that lets you specify them. Make sure the settings in that dialog accurately reflect your printer's installed accessories so you can take full advantage of them.
    See also
    Connecting to an IP Printing network printer
    I can't print to a network printer
    Good Luck
    SF

  • Running 10.8 drivers on 10.7

    Hello there,
    We have a Mini Server running OSX 10.8 and 25 iMacs running on 10.7 that are all bound through Windows AD and OpenDirectory.
    We have purchased a printer that has both 10.7 and 10.8 drivers available, I installed the printer to the server with the 10.8 drivers and then allocated it to each client through Workgroup Manager.
    The printer works great if you print directly from the MiniServer (running 10.8)
    However, when each iMac client tries to print (running 10.7) the image comes out in back and white. I am guessing this is because of the OSX version being slightly older that the, one on he server...
    Is there a solution for his? Obviously the 10.7 driver refuses to install on the 10.8 server.
    What are my options apart from updating each iMac to 10.8?
    Thanks for your time!
    iMac, Mac OS X (10.7.2)

    Hi, I'd uspect Page Setup or Printer settings the the Print Dailog, but of course no way to test here.

  • How to Verify Junk Filter Learning?

    Hi,
    With some helpful posts on this board, I setup Junk Mail filtering in 10.4 a few months ago. While it appears that everything is running fine, the system doesn't seem to learn much about what it should or shouldn't tag as spam. I'm using Spamtrainer and I did the Symlink fix. Is there a way to tell if the system is still learning? I receive this email each night and the numbers continue to go up...
    Learning SPAM...
    Learned new SPAM (junk mail)
    Learning HAM...
    Learned new HAM (not junk mail)
    Syncing SpamAssassin Database
    Displaying SpamAssassin Database Stats
    0.000 0 47298 0 non-token data: ham
    0.000 0 2062 0 non-token data: spam
    Deleting learned HAM and SPAM
    Done!
    Thanks.

    If you are already using the junkmail and notjunkmail accounts to train SpamAssassin, and the SpamTrainer script is reporting that the database is getting bigger and bigger, your spam filter is "functioning".
    The question is, "is it working for YOU"?
    The only qualitative measure is to compare the number of tagged and non-tagged spams.
    I use SpamSieve with Mail.app and its filter has become pretty smart about which messages I consider to be spam. It sets aside all the spam including those tagged as "**SPAM**" by SpamAssassin and those that SpamAssassin missed into a separate folder. Then I feed the SA non-tagged spam into the junkmail account. This way SA can learn from SS's expertise.
    Extended Keyboard Server   Mac OS X (10.4.7)  

  • URGENT Help Please! Cyrus database has disappeared

    Hey Guys & Gals--
    I've got a client whose mail system I was working on quite a bit yesterday to add Razor, update Spamassassin, etc. Late last night, all of the mail just disappeared. I had completed all my upgrades and was looking through headers of spam in a user's webmail to see how the changes were affecting incoming mail. I noticed a lot of BAYES_00 hits on spam, so the next thing I tried was to run spamtrainer and backup, then delete the bayes database to start fresh. I don't remember how long after that it was, but next thing I know, that user's webmail has ZERO messages. I have a look at the other accounts, and they're not even there. The only cyrus folders that show up are for the 3 users I logged into webmail as-- the rest have vanished!
    I've been at a funeral this morning, so I'm just now getting back on this. I'm gonna look around some more, but can someone please give me some ideas? I can likely restore from Retrospect backup, but that might be a day or 2 old. This is very strange. Please let me know what add'l info I can provide:
    Mac OS X Server v10.5.8
    Power Mac G5/2.3GHz DC
    2.5GB/500GB
    Spamassassin 3.2.5
    Razor 2.84 (I think-- the latest)
    latest Spamtrainer
    Getting these errors:
    Dec 14 15:00:07 HL-G5-Server postfix/trivial-rewrite[9883]: fatal: open database /etc/postfix/virtual_domains.db: No such file or directory
    Dec 14 15:00:07 HL-G5-Server postfix/smtpd[9850]: warning: connect #9 to subsystem private/rewrite: Connection refused
    Dec 14 15:00:08 HL-G5-Server postfix/master[39]: warning: process /usr/libexec/postfix/trivial-rewrite pid 9883 exit status 1
    Dec 14 15:00:08 HL-G5-Server postfix/master[39]: warning: /usr/libexec/postfix/trivial-rewrite: bad command startup -- throttling
    Dec 14 15:00:08 HL-G5-Server postfix/smtpd[9887]: warning: connect #1 to subsystem private/rewrite: Connection refused
    Thanks,
    Fred

    Ouch...no replies and only 18 views...guess these forums are too segmented and not "well traveled" enough. Anyway, big relief earlier-- got the system back online. And I have one whale of a Server Admin F* UP to tell you about.
    I had briefly forgotten that this server's mailstore is actually at /mailstore/spool/imap, rather than /var/spool/imap. And apparently, so had it. Since I maintain this particular server along w/ several others-- mine included-- I have them all in Server Admin, so that I can switch between them. I discovered today that SA had somehow overwritten this client's server w/ the settings for MY server!! ***?!?!?! It literally lifted the settings from my server's Mail component, and saved them to this client's server. Everything (in the Mail section) had been changed-- domain name, host name, relay via, filtering settings, and most importantly: mailstore location. So, cyrus had reverted back to the default IMAP mailstore location, but even that wasn't working, obviously due to discrepancies between that and the rest of the config files. Mail wasn't being delivered, and the accounts that you looked at were empty.
    What a bizarre problem! If I were to file a bug report, I'm not even sure what it'd say. I wouldn't even know how to reproduce it-- it didn't happen in any way that was visible to me at the time. Thankfully, only the Mail service suffered this overwriting...the rest of the service configurations were unharmed. I'd tell you to beware of SA doing this, but it may never happen again. Just strange...very strange...
    FT

Maybe you are looking for