Mail service appears to be corrupt

Mail service on my server appear to be corrupt. I had a crash about a week ago and periodic unresponsiveness of the server only occurs when mail services are enabled. The "Configure Mail service" button is appearing in the lower right hand corner now even though it is configured. I decided to repeat the process of configuring it using that button and all the processes attempt to start but mailing lists attempted and failed. That suggests there is something wrong. Should I try reconfiguring mail services from scratch. If so, how do I do it?

Things are still limping along. Digging into the log files, I find the following:
Here are some examples of error messages I see:
Jun 22 08:23:01 mini postfix/postfix-script[1442]: fatal: the Postfix mail system is not running
Jun 22 08:23:12 mini postfix/postfix-script[1486]: warning: damaged message: corrupt/1B1B3EA0C70
Jun 22 08:23:12 mini postfix/postfix-script[1487]: warning: damaged message: corrupt/1E360755445
Mailaccess.log
Jun 22 09:02:11 mini imaps[1564]: Fatal error: tlsstartservertls() failed
Jun 22 09:02:17 mini pop3s[1998]: Fatal error: tlsstartservertls() failed
I am considering doing a clean install of the entire server, but I do not want to do that unless I have to. Any help would be appreciated!

Similar Messages

Adobe e-mail service - anyone able to send photos to Facebook?

Facebook allows photos to be uploaded via a email address. None of the photos I send using Adobe E-Mail Service appear to arrive in Facebook. I am able to use the Adobe E-Mail client to send to other addresses without any problems and I can use other email clients to send to the Facebook address without any problems. Has anyone been able to get this to work for them?

Try sending the photos to your other email accounts(google, yahoo etc) and see if you get the email. This way we can figure out the cause of the issue.
~Sourabh

Mail service keeps corrupting.

Hello all,
My mail service keeps corrupting on my PowerMac running Mac OS X 10.3 Server.
I have reinstalled my server 3 times due to this problem.
Any one else having the same problem?
any solutions?
Thanks,
Chris

In addition to Alex's request for logs, and more info- I would point out that you are TWO major revs behind. I bet the downtime is worth more than the 499 upgrade (the 10 user limit applies only to filesharing). Give tiger or 10.5.1 a try.
It rarely makes sense to troubleshoot something thats already been fixed by Apple.
my $.02
Sean

Mail service does not appear on Webcenter Portal Services page.

Hi,
I did the connection to the mail service. But mail service does not appear. why this happen?

Hi,
I found where mail taskflow is but I didn't add. because my mail server "ExternalMail" dose not appear on the mail preferences page. I have any idea why? I restarted a few times. I am sory to have wasted your times. I am not WCP specialist. But I have to do.
Mail service connected. Test connection successful.

How do I set up Mail service for simple IMAP with SSL?

This should be easy but something isn't working. I have an Xserve with 10.6. I've enabled the Mail service, and want to add this server as an Account under Entourage. I've put in the IP address, my username/password, etc. but something isn't working. It appears under the Accounts folder tree, but when I try to send an email, I have a window that says "4.7.0 TLS not available due to local problem", Error -17099". I've also seen errors telling me the password is being sent in cleartext and I can't see where I enable the same (SSL, password needed) settings on the client and the server menus. Can someone clue me in - what are the right settings on both ends to have a simple IMAP w/o cleartext passwords? I had this working fine in 10.4 Server... Thanks!
Mike

I see dropdown menus for SMTP SSL and IMAP/POP SSL. I have them both set as "Require"
Do not to this if your server is handling incoming mail from outside your network.
Few remote mail servers will use SSL, so if your server requires SSL then you will not be able to receive mail from outside your domain.
'Use', on the other hand, advertises the fact your server supports SSL, so clients and remote servers can use if if they like, but this still allows other mail in.
What should "Authentication" be - "Login" or "Plain"
Do not use Plain unless you absolutely trust all your clients (i.e. they're all on a LAN or all using SSL). Plain sends an unencrypted password over the wire, which is susceptible to sniffing.
what will the client be using?
Whatever you set it to.
The client complains about a correct root certificate not being installed).
Are you using a self-signed certificate? if so, you're going to get that unless you do a lot of extra work. Getting a real certificate from a trusted authority should resolve this.
As for the slow connections, the use of SSL shouldn't significantly impact connection or delivery/transfer times, so something else is amiss.

Error while executing the service request : : Login to Mail Service failed

I am getting the following error in Convergence while attempting to authenticate against the mail service. I am able to access the backend mail service directly from a mail client with the same account without any issue.
Looking at the log file info:
==> iwc.log <==
AUTH: WARN from com.sun.comms.client.protocol.delegate.agent.LoginContextAgent Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:21,744 - Subject not found in session, creating one
AUTH: INFO from com.sun.comms.client.security.auth.modules.impl.SunAuthCallBack Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:21,745 - User LoginID is [email protected]
AUTH: INFO from com.sun.comms.client.security.auth.modules.impl.SunAuthCallBack Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:21,745 - User domain is mydomain.com
AUTH: INFO from com.sun.comms.client.security.auth.modules.impl.SunLDAPLoginModule Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:21,746 - Loaded UG LDAP pool...
AUTH: INFO from com.sun.comms.client.security.auth.modules.impl.SunLDAPLoginModule Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:21,750 - SunLDAPLoginModule:User [email protected] Authenticated
AUTH: INFO from com.sun.comms.client.security.auth.CommsUserInitContext Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:21,752 - Loading user entry from LDAP
AUTH: INFO from com.sun.comms.client.security.auth.CommsUserInitContext Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:21,754 - User entry loaded successfully
PROTOCOL: INFO from com.sun.comms.client.entity.user.sun.CommsUser Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:21,757 - IM Service is not enabled for user: sunImUser object class is not present
AUTH: INFO from com.sun.comms.client.web.IwcCookieManager Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:21,758 - Preferred language for user is en
PROTOCOL: WARN from com.sun.comms.client.protocol.delegate.UserPrefsCommandDelegate Thread httpSSLWorkerThread-8080-2 at 2009-03-10 04:33:22,404 - get_allprefs.iwc : Service is not enabled : im
PROTOCOL: INFO from com.sun.comms.client.entity.user.sun.CommsUser Thread httpSSLWorkerThread-8080-2 at 2009-03-10 04:33:22,405 - SMIME Service is not enabled for server
PROTOCOL: WARN from com.sun.comms.client.protocol.delegate.UserPrefsCommandDelegate Thread httpSSLWorkerThread-8080-2 at 2009-03-10 04:33:22,406 - get_allprefs.iwc : Service is not enabled : smime
PROTOCOL: WARN from com.sun.comms.client.protocol.delegate.agent.ClientOptionsAgent Thread httpSSLWorkerThread-8080-2 at 2009-03-10 04:33:22,407 - client preferences not found for domain: mydomain.com
PROTOCOL: INFO from com.sun.comms.client.entity.user.sun.CommsUser Thread httpSSLWorkerThread-8080-2 at 2009-03-10 04:33:22,418 - Client customization service is disbaled for the deployment
ADDRESS_BOOK: INFO from com.sun.comms.client.ab.coresrv.DBHandler Thread httpSSLWorkerThread-8080-1 at 2009-03-10 04:33:23,131 - Got DBPluginMap
PROXY_MAIL: INFO from com.sun.comms.client.web.services.sun.MailServiceProxy Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:23,139 - Performing Sun Mail Authentication
==> ../../messaging/log/http <==
[10/Mar/2009:04:33:23 -0400] fe-msg httpd[17512]: Account Information: connect [10.10.10.229:40786]
[10/Mar/2009:04:33:23 -0400] fe-msg httpd[17512]: General Information: [10.10.10.229:40786] POST /login.msc HTTP/1.1
[10/Mar/2009:04:33:23 -0400] fe-msg httpd[17512]: Account Notice: badlogin: [10.10.10.229:40786] proxy admin [email protected] domain=us.mydomain.com: user not found
==> iwc.log <==
PROXY_MAIL: ERROR from com.sun.comms.client.web.services.sun.MailServiceProxy Thread httpSSLWorkerThread-8080-0 at 2009-03-10 04:33:25,160 - cookies not present from MS response
==> ../../messaging/log/http <==
[10/Mar/2009:04:33:25 -0400] fe-msg httpd[17512]: Account Notice: close [10.10.10.229:40786] [unauthenticated] 2009/3/10 4:33:23 0:00:02 248 0 0
It appears that the authentication mechanism is trying to authenticate against the domain "us.mydomain.com" rather than "mydomain.com". At one point I added "us.mydomain.com" as an alias to the domain "mydomain.com" but it has since been removed. I am unsure where Convergence is pulling this information from. Any advice would be appreciated.

highamjd1 wrote:
[10/Mar/2009:04:33:23 -0400] fe-msg httpd[17512]: Account Information: connect [10.10.10.229:40786]
[10/Mar/2009:04:33:23 -0400] fe-msg httpd[17512]: General Information: [10.10.10.229:40786] POST /login.msc HTTP/1.1
[10/Mar/2009:04:33:23 -0400] fe-msg httpd[17512]: Account Notice: badlogin: [10.10.10.229:40786] proxy admin [email protected] domain=us.mydomain.com: user not found
It appears that the authentication mechanism is trying to authenticate against the domain "us.mydomain.com" rather than "mydomain.com". At one point I added "us.mydomain.com" as an alias to the domain "mydomain.com" but it has since been removed. I am unsure where Convergence is pulling this information from. Convergence determines the default domain from the URL that you used to connect e.g. if you used "http://myserver.us.mydomain.com/iwc" it sets the default domain to "us.mydomain.com"
http://wikis.sun.com/display/CommSuite/Sun+Convergence+Administrative+Tasks#SunConvergenceAdministrativeTasks-HowdoIcustomizetheLoginpagebasedonthedomainnameintheURLtoaccesstheConvergenceclient%3F
Rather then use the default domain for the proxy administrator, you can hard-code the domain e.g.
./iwcadmin -w <password> -o mail.proxyadminid -v [email protected]
./iwcadmin -w <password> -o cal.proxyadminid -v [email protected] need to restart the application-server domain hosting Convergence for the changes to become active.
Regards,
Shane.

PSE 7.0 Crashes w/Adobe E-Mail Service

PSE 7.0 crashes with the following response when I try to send an image using Photomail or as an attachment.
It appears to work fine (or at least gets to the point where I can edit the message) when using Outlook or Outlook Express. However, we use Thunderbird, so would prefer to use the Adobe E-Mail service. We upgraded from PSE 4, which worked on the Adobe E-Mail Service with no issues.
"Adobe Photoshop Elements 7.0 (Organizer) has encountered a problem and needs to close. We are sorry for the inconvenience."
It gets to the point where the email shows up on the right-hand sidebar, then the mouse goes to an hourglass for a few seconds and then the above error message.
I'm running XP Pro SP3.
Any ideas?
Thanks,
Marc

A couple of other thoughts:
1. Perhaps PSE 7 doesn't like the PSE 4 address book. That happens occasionally, though I recall it caused other symptoms, not a crash. You could find all copies of the file "contacts.xml" and delete them, and see if that helps. Different versions of PSE can hide these in dfferent folders, so configure Windows Explorer to show hidden files and then search your entire C drive.
2. You could try the sledgehammer approach of uninstalling PSE 7, manually removing all traces left behind by the uninstaller, and then re-installing in what Adobe calls "simplified mode", with anti-virus and all other background services disabled:
http://www.johnrellis.com/psedbtool/photoshopelements-6-7-faq.htm#Try_Adobes_generic_troub leshooting

Cannot edit text sufficiently within the Photo E-mail service (Elements 11)

Greetings to all, a little help required please....
Just purchased Elements 11 - excellent product at the price and am having a great deal of fun using it.
One frustrating flaw within the Photo E-mail service....
I've been using the Adobe e-mail service and it sends e-mail just fine having set up the account as required.
Each new e-mail appears with the text "Here are the files I want to share with you."
Now I'd like to replace that with some different text. I can delete it and replace it on the first editing page, and it stays there after I click 'Next' to the following 'Customise Layout' section.
Clicking on 'Next' after this page, we have the final e-mail page, and the text has reverted back to 'Here are the files I want to share with you' again......
I can delete it again, but cannot type in any further text anywhere on that screen.
Any help on this??
Shame, because the programme generally is superb. Have transformed a number of pics already, and it would be nice to use the e-mail prog that's built in as I'd like to use it, adding text as required. Otherwise, why have all the font and layout choices available if not??

I tried the same and its changing at my side.
Try to reset you preferences :
Open PSE just to the Welcome Screen and press Ctrl + Alt + Shift (Mac: Cmd + Opt + Shift) and keep pressing.
Click on organizer from the Welcome Screen but keep pressing the keyboard keys - don't let up.
You will get a little dialog box asking if you want to delete the settings file. Click "Yes." You can let go of the keyboard keys now.
This will rebuild the file for you and often solves the issue.
Try to send the email again.
Thanks
Harshit

Mail Services and Backup MX/mail services

I have a question regarding backup mail services and acceptance of incoming messages.
As of last week, I am new to the world of Mail Services in OS X Server 10.4. Last week, our current email server died and I was forced to take an Xserve G4 that had OS X Server 10.4.10 (unlimited) and setup mail services in order to receive and send mail. I went to a website called osx.topicdesk.com and followed their instructions for front-line spam defense and clamAV updating (and switching to clamd) which all appears to be working OK.
Because our old email server was having some hardware issues for the past several months that I had been working on, I had the feeling that failure might happen. So what I did a few months ago was sign up for Google Apps and their mail service. I added their MX records under my primary mail server so that in the case of a failure of my server, all mail would be passed to Gmail and everyone could get to their mail immediately (as any prolonged downtime won't work since this is a daily newspaper). This did work, for when my old server died, all the mail went immediately to the appropriate Google Apps mail accounts. After about a day of setup on the Xserve, I got the mail services running OK. I gave it all the same IP info as the previous server so that the mail would just start working without having to edit DNS records and such.
The problem I am having seems to be this: While I get most mail, some of my mail is still getting sent to the Google Apps accounts. For example, I have one user, Jane Doe, at [email protected] who gets most of their mail from connecting to the OS X Server, but about 2 - 3 emails a day get passed over to the Google Apps server. I have run DNSStuff.com tests on the mail connections, and I always get a good fast response out of the OS X server. My thinking was that mail would only be passed over to backup mail servers if the first server listed in the MX records was unavailable or offline. In all of my testing, my server has responded every time. The server is connecting via a gigabit network connection (good response), is a dual G4 1.33Ghz, 60GB hard drive with 2GB of RAM. This server doesn't do anything else except run Mail services (and web services for squirrelmail), it doesn't even have file sharing turned on.
My old email server was a Pentium II 400mhz with 348MB of RAM, and it caught all mail, so I know that the Xserve should be able to handle the load without being busy. Is there any reason that my Xserve would be rejecting mail and they would be siphoned off to the Google accounts?
Is this even a possible scenario to fix the problem: Setup a backup MX "queue" account somewhere with one of the BackupMX online services that will catch these stray emails and then send it back to the primary server? But then can be turned off quickly to allow mail to go to the Google Apps accounts if my server went down. In other words, the DNS record would show my primary as 1, the backup mx service as 2, the google servers as 3, 4, 5 and so on. Can Backup MX services be turned off quickly like that so that if my server is down and I need email to be accessed immediately then just "turn off" the service (without modifying the DNS since that will take too long to propogate) so that mail bounces over to Google immediately?
I guess my first troubleshooting attempt should be to figure out why my OS X Server is not accepting all mail.
Thanks in advance for any help and suggestions.
G

When I first started with the company about 2 years ago, they had a really old Linux box (about 10 years old) that was used to host email. It was horrible. It would run OK, but if it ever went down or had a hiccup, it was an all-day project trying to deal with it. Unfortunately, the company did not want to spend any money on a new server or software of course. The answer I got was "well, it's just email". Sure, but if email EVER goes down for more than 15 seconds, all of the employees go into seizures and scream that the place will have to shut down. I work for a daily newspaper, so everything is running 24/7.
Knowing that one day that box was going to just die, I looked into some backup MX services. As pterobyte had mentioned is that with a secondary MX, the problem is that no one can get their mail at all until the primary server comes back up. I knew if that server ever went down, I had nothing to take its place, so I would have to order something or get a server from another location to use. Being unfamiliar with OS X Mail Services, I knew it would take me awhile to try to setup all the users and such. All the while at the newspaper, they would be screaming bloody murder that they needed email working for news, ads coming in, etc. And when I say if it goes down for 10 minutes they start getting into fetal positions in the floor screaming, I'm not far off exaggerating. I figured I was going to need something in place for them to check email while I worked on the primary.
Back in September/October 2007, I discovered that Google hosted free mail services call Google Apps. With Google, I could setup "matching" accounts and would still be able to use my domain name. You just had to use Google's servers as the higher numbered MX records. Luckily, I set up in time for back in November, the Linux box finally died. When it did, all email successfully went to their counterparts on the Google's servers. I had everyone go to the special Google Apps login site and they were able to check their email fine. I ended up having to take a box from another site and then I had another copy of OS X Server 10.4, so I loaded it and successfully setup email (with lots of help from pterobyte's tutorials). They were on Google's systems for about 2 days.
After the disaster, I went around and setup IMAP accounts on everyone's Thunderbird clients so that if our email server went down, they just simply check that mailbox. The nice thing with the Google IMAP is that they automatically see if emails are hitting that inbox from within their client. Right now, it still remains that the 2 original companies that hit the Google accounts still do. The city government email and email from a sister newspaper. Everything else gets directed to the primary server. Although sometimes that doesn't even work, as tonight I am having to work on a problem involving the Clamd app acting up - taking up nearly 100% of the CPU which in turn caused all the mail to be hung up in the queue. So since the mail server didn't "reject" the mail and grabbed it, their gmail "backup" didn't get anything in it. It was taking me forever to figure out what was going on, so I almost just told them to unplug the network cable from the email server so that everything definitely would be forwarded to gmail.
My wish is that I had a secondary mail server in house to not even worry about it - or better yet, I wished I could just outsource the email administration itself to someone else (like webmail.us which is now MailSafe). That way for the most part, they wouldn't lose email. But of course, the company doesn't want to spend the money or invest. Just gripe when the older equipment we have won't support what they want to do with it.
But the story above explains why I did what I did with the Google's servers. I do appreciate everyone's help and input on this particular problem. I am going to look around at the DNS and talk to at least our sister newspaper to see if I can see how they have their DNS setup.
Thanks!
G

Mail service rejecting all external mail

I am running Snow Leopard Server 10.6.2 and when I began setting up the mail services I was able to get internal mail working just fine, however, any messages sent from external sources are rejected. The sender receives a permanent delivery failure message: The error that the other server returned was: 550 550 5.1.1 <emailaddress>: Recipient address rejected: undeliverable address: User unknown (state 14). I should note that I am using DynDNS's Mailhop Relay service, but that appears to be working normally as I can see the connections from the mailhop servers in my SMTP log.
I have tried disabling greylisting as posted in other threads to no effect. I have disabled filtering of any kind, still to no effect. At this point, I'm not sure what else to try, as at this point I'm in a little over my head.
Any assistance would be most appreciated, thanks.

I finally figured it out. My SMTP security settings were being mis-reported by Server Admin. The server was requiring authentication from approved mail servers when it wasn't supposed to.

Mail Services: Is fetchmail supported?

I have a Mac Pro running OS X and is configured to provide mail and DNS services. When I need to use a VPN connection, these services become unavailable to other systems on my network.
As a solution to this problem a purchased a Mac Mini Server. After reading the Mail Services documentation, it appears that fetchmail might not be provided as part of the service. Is fetchmail supported?
Currently, I use fetchmail to download mail from my ISP's mail server using POP. Systems on my local network use IMAP to access mail on my Mac Pro.

I have a Mac Pro running OS X and is configured to provide mail and DNS services. When I need to use a VPN connection, these services become unavailable to other systems on my network.
Without knowing more detail I'd be inclined to say your network/VPN is misconfigured…
As a solution to this problem a purchased a Mac Mini Server.
That's one way of fixing it, I suppose
After reading the Mail Services documentation, it appears that fetchmail might not be provided as part of the service. Is fetchmail supported?
That depends on what your definition of 'support' is. Server Admin doesn't 'support' it (you can't use Server Admin to configure fetchmail), and Apple don't 'support' it (you can't call AppleCare with problems configuring/using it), but that doesn't mean it doesn't run, or that you can't configure it to do what you want.
In other words, fetchmail is installed, can be used, but you're on your own as far as configuring/maintaining it is concerned.

Mail services 10.4.10 refuses to work.

Hello,
I'm a newbie to OS X, so please help - I need to get
a local eMail system going here in the office, and it
just doesn't want to work.
I've been to every site imagineable searching for answers
on how to configure OS X Mail Services, but it still
doesn't work.
DNS:
- I have DNS running on this server,
- I've checked "This machine is a mail server for the zone" for the server,
- I've added aliases "pop3", "smtp", and "mail" for the server.
User accounts:
- Open Directory seems to be running, although "start service" is greyed-out,
- this is an "Open Directory Master",
- I've checked "Enable Secure Sockets Layer" on the Protocols tab,
- I've checked all the encryption options in the Policy->Security tab.
Workgroup Manager:
I followed some tutorial to create the directory "/LDAPv3/127.0.0.1",
and added the user "dspratling" to it, which DOES show up in the
Mail Services automagically.
I'm only looking to create Mail Users ("dspratling" for starters),
but trying everything, I created a Home directory for this user,
since I was getting error messages in the LDAP log saying: "... object
class 'posixAccount' requires attribute 'homeDirectory'.
- I have checked the "Mail: Enabled" under the Mail tab,
- I checked "POP only" under Mail Access,
- I put "mail.mme.lan" for the Mail Server.
- Under the Advanced tab, I have User Password Type selected as "Open Directory".
Error Messages appearing in the logs:
Other error messages in LDAP log include:
"slapd[59]: SASL [conn=3740] Failure: GSSAPI Error: Miscellaneous failure (No principal in keytab matches desired name)"
kdc Log has:
"xserv.mme.lan krb5kdc[296](info): AS_REQ (7 etypes {18 17 16 23 1 3 2}) 10.0.0.9: ISSUE authtime 1195081950, etypes {rep=16 tkt=16 ses=16}, [email protected] for krbtgt/[email protected]"
kadmin Log has:
"xserv.mme.lan kadmin.local[22270](info): No dictionary file specified, continuing without one."
however, good news in the "Password Service Server Log":
"RSAVALIDATE: success"
and
"AUTH2: {0x000000..., diradmin} CRAM-MD5 authentication succeeded."
Now, in the Mail logs...
Mail Access log has:
"xserv master[23575]: empty option value on line 19 of configuration file"
and
"xserv master[23575]: exiting"
POP log has:
"xserv pop3[23444]: AOD: getting mail attribute for user: dspratling
"xserv pop3[23444]: AOD: mail enabled"
"xserv pop3[23444]: AOD: mbrcheck_servicemembership with accessallservices"
"xserv pop3[23444]: AOD: checkServiceACL: Mail SACL not enabled"
From the Client-side (MacMini 10.4.10), it reports "The server "mail.mme.lan" refused to allow a connection on port 995."
All other network-based services work fine with the server, even in our
heterogenous PC network - only the Mail Service is not working.
Please - ANY SUGGESTIONS WOULD BE APPRECIATED
Thanks,
Dylan

Brett -
Thanks for the suggestions. This is not an internet-facing
server, however. Strictly intra-office eMail is all I'm after
for now. The DNS works well, resolving "mail.mme.lan" to the
correct IP (via the CL "host" command), and the CL "named-checkzone"
command returned OK as well. I haven't changed the DNS for weeks,
so I'm quite sure every cache has been refreshed.
I'm going with "Plan B", until this can be resolved ...
And I know it's pure evil, but: "Parallels -> Win2K -> mailEnable"
I got it configured and working in minutes. (Parallels was already
running on this server, serving printers, and mailEnable was free
Would still rather not depend on MS.
Thanks,
Dylan

Mail Services Tutorial

Does anyone know of a good place to get a tutorial on how to use mail services? I've used the Mac OS X help pages that come with server admin, and that has just left me confused. A tutorial that walks you through everything step by step would be great. Thank you.

OK, I think I've found a clue.
When 10.6 server was released, there was an Apple article to fix some log messages that were occurring. Here is an excerpt:
*The Mail Server mail.log may contain encryption warnings after upgrading to Mac OS X Server v10.6.
*For example, warnings such as these may appear in the mail.log:
* warning: no entropy source specified with parameter tlsrandomsource
* warning: encryption keys etc. may be predictable
*For improved encryption in Mac OS X Server v10.6, and to avoid this issue, open Terminal and execute this *command:
*sudo postconf -e tlsrandomsource=/dev/urandom
I've found that if you comment out this line in main.cf:
tlsrandomsource = /dev/urandom (at the very end of the file for me)
then TLS will start up OK. It will display the warning messages in the apple article above, but it starts up and handles TLS !!
Next step is to figure out what broke in the random source and fix the encryption warnings. Anyone want to take a stab at this?

DispCache appears to be corrupt. getDispFile() returned null

Hi all,
I am deploying web services in production mode on Solaris 8 and Win2K. I get identical
errors when I invoke the web service from a Java client: "DispCache appears to
be corrupt. getDispFile() returned null"
The complete stack trace is posted below.
The web services work perfectly fine when I use Workshop to deploy within its
environment.
WebLogic version 7.0.4.0. Used jwsCompile tool separately on Win2K and Solaris
to package the ear file to remove any platform dependencies (I noticed while debugging
the problem that the EJBs generated by the jwsCompile hardcodes the ProjectPath
and LoaderPath in the deployment descriptor, weird and totally unncessary).
Any help is highly appreciated.
Once thing I noticed that WebLogic's web service solution is overly complex to
deploy in production once developed using Workshop. There is no way in hell I
can debug this problem, hey it never occurs in the Workshop and there is no other
debug environment!
Thanks, Sanjay ----------------------------------------------------------------
The stack trace (identical from Solaris and Win2K):
SERVICE FAULT: Code: SOAP-ENV:Server String: DispCache appears to be corrupt.
getDispFile() returned null. Detail: END SERVICE FAULTweblogic.jws.control.ServiceControlException:
SERVICE FAULT: Code: SOAP-ENV:Server String: DispCache appears to be corrupt.
getDispFile() returned null. Detail: END SERVICE FAULT at weblogic.knex.control.ServiceControlImpl.invoke(ServiceControlImpl.java:315)
at weblogic.knex.control.ControlHandler.invoke(ControlHandler.java:316) at $Proxy123.GetUserAttributes(Unknown
Source) at EWSWebServiceTest.testGetUserAttributes(EWSWebServiceTest.java:58)
at java.lang.reflect.Method.invoke(Native Method) at weblogic.knex.dispatcher.DispMethod.invoke(DispMethod.java:211)
at weblogic.knex.dispatcher.Invocable.invoke(Invocable.java:454) at weblogic.knex.bean.WebBean.invokeBase(WebBean.java:195)
at weblogic.knex.bean.WebServiceBean.invoke(WebServiceBean.java:70) at weblogic.knex.bean.WebServiceBean_v4x0yn_EOImpl.invoke(WebServiceBean_v4x0yn_EOImpl.java:46)
at weblogic.knex.bean.WebDispatcherBean.invoke(WebDispatcherBean.java:103) at
weblogic.knex.bean.RemoteDispatcherBean.invoke(RemoteDispatcherBean.java:127)
at weblogic.knex.bean.RemoteDispatcherBean_2c75i3_EOImpl.invoke(RemoteDispatcherBean_2c75i3_EOImpl.java:46)
at weblogic.knex.bean.RemoteDispatcherBean_2c75i3_EOImpl_WLSkel.invoke(Unknown
Source) at weblogic.rmi.internal.ServerRequest.sendReceive(ServerRequest.java:159)
at weblogic.rmi.cluster.ReplicaAwareRemoteRef.invoke(ReplicaAwareRemoteRef.java:263)
at weblogic.rmi.cluster.ReplicaAwareRemoteRef.invoke(ReplicaAwareRemoteRef.java:230)
at weblogic.knex.bean.RemoteDispatcherBean_2c75i3_EOImpl_WLStub.invoke(Unknown
Source) at weblogic.knex.dispatcher.Dispatcher.remoteDispatch(Dispatcher.java:120)
at weblogic.knex.dispatcher.Dispatcher.dispatch(Dispatcher.java:66) at weblogic.knex.dispatcher.HttpServer.exploreExec(HttpServer.java:458)
at weblogic.knex.dispatcher.HttpServer.doGet(HttpServer.java:364) at javax.servlet.http.HttpServlet.service(HttpServlet.java:740)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:853) at weblogic.servlet.internal.ServletStubImpl$ServletInvocationAction.run(ServletStubImpl.java:1058)
at weblogic.servlet.internal.ServletStubImpl.invokeServlet(ServletStubImpl.java:401)
at weblogic.servlet.internal.ServletStubImpl.invokeServlet(ServletStubImpl.java:306)
at weblogic.servlet.internal.WebAppServletContext$ServletInvocationAction.run(WebAppServletContext.java:5445)
at weblogic.security.service.SecurityServiceManager.runAs(SecurityServiceManager.java:780)
at weblogic.servlet.internal.WebAppServletContext.invokeServlet(WebAppServletContext.java:3105)
at weblogic.servlet.internal.ServletRequestImpl.execute(ServletRequestImpl.java:2588)
at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:213) at weblogic.kernel.ExecuteThread.run(ExecuteThread.java:189)

Sanjay,
Would it be possible for you to send us your application, so that we can
reproduce the error here. Let me know if you are unable to post on a public
newsgroup, and we can take it offline.
I would also suggest you to try using version 8.1, because it offers loads
of features that make web services development a much more richer and easier
experience.
Regards,
Anurag
"Sanjay" <[email protected]> wrote in message
news:[email protected]...
>
Hi all,
I am deploying web services in production mode on Solaris 8 and Win2K. Iget identical
errors when I invoke the web service from a Java client: "DispCacheappears to
be corrupt. getDispFile() returned null"
The complete stack trace is posted below.
The web services work perfectly fine when I use Workshop to deploy withinits
environment.
WebLogic version 7.0.4.0. Used jwsCompile tool separately on Win2K andSolaris
to package the ear file to remove any platform dependencies (I noticedwhile debugging
the problem that the EJBs generated by the jwsCompile hardcodes theProjectPath
and LoaderPath in the deployment descriptor, weird and totallyunncessary).
>
Any help is highly appreciated.
Once thing I noticed that WebLogic's web service solution is overlycomplex to
deploy in production once developed using Workshop. There is no way inhell I
can debug this problem, hey it never occurs in the Workshop and there isno other
debug environment!
Thanks,Sanjay ----------------------------------------------------------------
The stack trace (identical from Solaris and Win2K):
SERVICE FAULT: Code: SOAP-ENV:Server String: DispCache appears to becorrupt.
getDispFile() returned null. Detail: END SERVICEFAULTweblogic.jws.control.ServiceControlException:
SERVICE FAULT: Code: SOAP-ENV:Server String: DispCache appears to becorrupt.
getDispFile() returned null. Detail: END SERVICE FAULT atweblogic.knex.control.ServiceControlImpl.invoke(ServiceControlImpl.java:315)
at weblogic.knex.control.ControlHandler.invoke(ControlHandler.java:316) at$Proxy123.GetUserAttributes(Unknown
Source) atEWSWebServiceTest.testGetUserAttributes(EWSWebServiceTest.java:58)
at java.lang.reflect.Method.invoke(Native Method) atweblogic.knex.dispatcher.DispMethod.invoke(DispMethod.java:211)
at weblogic.knex.dispatcher.Invocable.invoke(Invocable.java:454) atweblogic.knex.bean.WebBean.invokeBase(WebBean.java:195)
at weblogic.knex.bean.WebServiceBean.invoke(WebServiceBean.java:70) atweblogic.knex.bean.WebServiceBean_v4x0yn_EOImpl.invoke(WebServiceBean_v4x0yn
_EOImpl.java:46)
at weblogic.knex.bean.WebDispatcherBean.invoke(WebDispatcherBean.java:103)at
>
weblogic.knex.bean.RemoteDispatcherBean.invoke(RemoteDispatcherBean.java:127
atweblogic.knex.bean.RemoteDispatcherBean_2c75i3_EOImpl.invoke(RemoteDispatche
rBean_2c75i3_EOImpl.java:46)
atweblogic.knex.bean.RemoteDispatcherBean_2c75i3_EOImpl_WLSkel.invoke(Unknown
Source) atweblogic.rmi.internal.ServerRequest.sendReceive(ServerRequest.java:159)
atweblogic.rmi.cluster.ReplicaAwareRemoteRef.invoke(ReplicaAwareRemoteRef.java
:263)
atweblogic.rmi.cluster.ReplicaAwareRemoteRef.invoke(ReplicaAwareRemoteRef.java
:230)
atweblogic.knex.bean.RemoteDispatcherBean_2c75i3_EOImpl_WLStub.invoke(Unknown
Source) atweblogic.knex.dispatcher.Dispatcher.remoteDispatch(Dispatcher.java:120)
at weblogic.knex.dispatcher.Dispatcher.dispatch(Dispatcher.java:66) atweblogic.knex.dispatcher.HttpServer.exploreExec(HttpServer.java:458)
at weblogic.knex.dispatcher.HttpServer.doGet(HttpServer.java:364) atjavax.servlet.http.HttpServlet.service(HttpServlet.java:740)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:853) atweblogic.servlet.internal.ServletStubImpl$ServletInvocationAction.run(Servle
tStubImpl.java:1058)
atweblogic.servlet.internal.ServletStubImpl.invokeServlet(ServletStubImpl.java
:401)
atweblogic.servlet.internal.ServletStubImpl.invokeServlet(ServletStubImpl.java
:306)
atweblogic.servlet.internal.WebAppServletContext$ServletInvocationAction.run(W
ebAppServletContext.java:5445)
atweblogic.security.service.SecurityServiceManager.runAs(SecurityServiceManage
r.java:780)
atweblogic.servlet.internal.WebAppServletContext.invokeServlet(WebAppServletCo
ntext.java:3105)
atweblogic.servlet.internal.ServletRequestImpl.execute(ServletRequestImpl.java
:2588)
at weblogic.kernel.ExecuteThread.execute(ExecuteThread.java:213) atweblogic.kernel.ExecuteThread.run(ExecuteThread.java:189)
>
>

Upgrading 10.3.9 - 10.4.3 has blown up Mail Services! Heeeeellllpppp!

I have to upgrade a customer's Xserve from 10.3.9 to 10.4 and I decided to upgrade my own server first - just to see what might go wrong. After running the installer and doing setup I ran software update and got to 10.4.3. Everything came up as it should, but Mail Service is broken. Going to Server Admin>Mail Services>Maintenance>Accounts each mail account shows 'unknown Mailbox Path'. Reconstructing the database doesn't help. Choosing an account and pressing reconstruct gives the message 'fatal error: can't read mailboxes file'.
I really don't want to try this on the customer's system with many more users and emails than we've got if I can't even get our system to run (not to mention we've got no mail service!)
Any help at all would be hugely appreciated.

I took a look at the doc but it doesn't seem to apply. The previous mail database was in the default location, and to my less-than-fully-trained eyes the files and folders appear to be where they should be now. The database tab under maintenance shows the locations of the database files, and their sizes. The sizes in my case seem to be right - i.e. the same as they were before. (Mind you, the 'big' upgrade I have to do next week the database is not in the default location so I'm grateful for that piece of info in any case.)
Email clients can't log in to mail services (connection timed out). Newly-arrived mail is showing in the 'mail queue' screen.
The IMAP log is repeatedly saying
skiplist:invalid magic header: /var/imap/mailboxes.db (also appearing in Mail Access log)
DBERROR: opening /var/imap/mailboxes.db: cyrusdb error (also appearing in Mail Access log)
Fatal error: can't read mailboxes file
service imap pid xxxxx in READY state: terminated abnormally
executed
The SMTP log is saying normal things.
Pushing the 'Repair' button under databases doesn't seem to do anything. The button just goes grey and then reappears after a few seconds. Under 10.3 there used to be a repair log you could look at. Is there an equivalent log being generated in 10.4 (and if so where is it!)?
I'm about to create a test user and see what happens with that.

Mail service appears to be corrupt

Similar Messages

Maybe you are looking for