Not the expert here so cant help directly, but the experts are going to need a bit more info..   Such as how you are sending, errors getting back, etc.   Too many possibilities just to guess.

Wed Jan 03 2024 08:10:50 EST from Virendra

Hello Team,
We have configured Citadel, we are successfully receiving mails  from outside but we can not send mail to outside email address.
Please guide what have to do for it

Thanks

Hi, after upgrading to the latest citadel version all my mail adreses are useless, i have been using 5 numbers in front of @

is there a way to re-enable the possibility? tested with 4 numbers, that is the max allowed ??

pleas help

-geir

Hello,

Citadel server Version 972 suddenly starts crashing. Last entry in the syslog show strange entries like this:

Jan  9 09:28:05 [host] citserver[2665335]: citserver[2665335]: crypto: SSL_accept failed: (null)
Jan  9 09:28:05 [host] citserver[2665335]: citserver[2665335]: sysdep: client_write(50 bytes) failed: Broken pipe
Jan  9 09:28:05 [host] citserver[2665335]: crypto: SSL_accept failed: (null)
Jan  9 09:28:05 [host] citserver[2665335]: citserver[2665335]: context: session 17576 (SMTPs-MTA) ended.
Jan  9 09:28:05 [host] citserver[2665335]: sysdep: client_write(50 bytes) failed: Broken pipe

Still operational for a couple of hours apparently, followed by these shortly before the initial crash:

Jan  9 12:19:23 [host] citserver[2665335]: msgbase: message #97535 was not found
Jan  9 12:19:23 [host] citserver[2665335]: smtpclient: 97535 does not exist
Jan  9 12:19:36 [host] citserver[2665335]: free(): invalid pointer
Jan  9 12:19:36 [host] systemd[1]: citadel.service: Main process exited, code=killed, status=6/ABRT
Jan  9 12:19:36 [host] systemd[1]: citadel.service: Failed with result 'signal'.
Jan  9 12:19:37 [host] systemd[1]: citadel.service: Scheduled restart job, restart counter is at 11.
Jan  9 12:19:37 [host] systemd[1]: Stopped Citadel Server.

From then on restarts end in something like this:

Jan  9 12:20:56 [host] citserver[2699682]: citserver[2699682]: fulltext: flushing index cache to disk (0% complete)
Jan  9 12:20:56 [host] citserver[2699682]: fulltext: flushing index cache to disk (0% complete)
Jan  9 12:20:56 [host] citserver[2699682]: citserver[2699682]: db: BDB0689 cdb.09 page 13525 is on free list with type 7
Jan  9 12:20:56 [host] citserver[2699682]: citserver[2699682]: db: BDB0061 PANIC: Invalid argument
Jan  9 12:20:56 [host] citserver[2699682]: citserver[2699682]: db: PANIC: Invalid argument
Jan  9 12:20:56 [host] citserver[2699682]: db: BDB0689 cdb.09 page 13525 is on free list with type 7
Jan  9 12:20:56 [host] citserver[2699682]: db: BDB0061 PANIC: Invalid argument
Jan  9 12:20:56 [host] citserver[2699682]: db: PANIC: Invalid argument
Jan  9 12:20:56 [host] systemd[1]: citadel.service: Main process exited, code=killed, status=6/ABRT
Jan  9 12:20:56 [host] systemd[1]: citadel.service: Failed with result 'signal'.
Jan  9 12:20:56 [host] systemd[1]: citadel.service: Scheduled restart job, restart counter is at 13.
Jan  9 12:20:56 [host] systemd[1]: Stopped Citadel Server.

And eventually looping this one getting restarted by systemd:

Jan  9 12:21:27 [host] citserver[2699722]: citserver[2699722]: msgbase: message #97535 was not found
Jan  9 12:21:27 [host] citserver[2699722]: citserver[2699722]: smtpclient: 97535 does not exist
Jan  9 12:21:27 [host] citserver[2699722]: msgbase: message #97535 was not found
Jan  9 12:21:27 [host] citserver[2699722]: smtpclient: 97535 does not exist
Jan  9 12:21:29 [host] citserver[2699722]: citserver[2699722]: db: cdb_fetch(9): BDB0075 DB_PAGE_NOTFOUND: Requested page not found
Jan  9 12:21:29 [host] citserver[2699722]: db: cdb_fetch(9): BDB0075 DB_PAGE_NOTFOUND: Requested page not found
Jan  9 12:21:29 [host] systemd[1]: citadel.service: Main process exited, code=killed, status=6/ABRT
Jan  9 12:21:29 [host] systemd[1]: citadel.service: Failed with result 'signal'.
Jan  9 12:21:29 [host] systemd[1]: citadel.service: Scheduled restart job, restart counter is at 25.
Jan  9 12:21:29 [host] systemd[1]: Stopped Citadel Server.

Thought database was somehow corrupted so replaced it with latest backup and afterwards was working fine again but lost some data obviously. Only thing is now that the error related to the invalid message id is still there and logged frequently:

Jan 10 09:02:56 [host] citserver[10806]: citserver[10806]: msgbase: message #97535 was not found
Jan 10 09:02:56 [host] citserver[10806]: citserver[10806]: smtpclient: 97535 does not exist
Jan 10 09:02:56 [host] citserver[10806]: msgbase: message #97535 was not found
Jan 10 09:02:56 [host] citserver[10806]: smtpclient: 97535 does not exist

So I am wondering whether anybody has an idea on two questions:

1. What could have caused the database corruption and is there a tool that can be used to check and restore database integrity without data loss?
2. What is causing the errors related to message #97535 and how can I fix this?

Any ideas are more than welcome.

Regards,

P.

Hello dear admin and support I hope you are in good mood today actually I found a important bug in citadel mail server and it could endanger sensitive data in local and global citadels I never ill-use that bug and it's information as you now no pain no gain I decided to inform you that report by email but can I expect prize ?(I am bug hunter and receive prize in trust wallet) your helping and cooperation can encourage me to trying find more bugs we can make better internet environment by the way I don't want too much money after observing and confirmation you can evaluate my request

so please send me email here my user is admin1234

and my email is     rainyboy95@yahoo.com

Best regards

lol

Hi what are you meaning?you wrote lol 

Who are you and why write this things?

First of all many thanks for the support - as always very quick and to the point.

1. Upgrade to Citadel Server v993 or later. We literally spent all of last summer improving the database layer, and it is now much more robust, even in the face of server crashes or other problems.

Done that and Citadel and DB are now version 996. Let's see whether we see again any DB issues. Had to restore DB from backup twice in the last two years.

2. You might have instructions in your outbound SMTP queue to deliver that message but the message itself doesn't exist. If you want to, you can go to the hidden room "__CitadelSMTPSpoolout__" and your queue will be in there.
If you find delivery instructions that reference that message, just delete them.

My understanding is only Aide can enter that room, right? I can get there two ways:
1. Go via Administration -> View the outbound SMTP queue
2. Go via Advanced -> Go to a 'hidden' room
Both ways no messages are shown:

Is there a way to remove the message number via command line client?

BTW, is there a documentation on error and return codes somewhere? I noticed that 'systemctl stop citadel' does not exit clean but with retun code 15. Not sure what this means.

Regards,

P.

Thu Jan 11 2024 15:13:25 EST from IGnatius T Foobar Subject: Re: Citadel Server crashes

1. What could have caused the database corruption and is there a tool that

can be used to check and restore database integrity without data loss?

2. What is causing the errors related to message #97535 and how can I fix

this?

1. Upgrade to Citadel Server v993 or later. We literally spent all of last summer improving the database layer, and it is now much more robust, even in the face of server crashes or other problems.

2. You might have instructions in your outbound SMTP queue to deliver that message but the message itself doesn't exist. If you want to, you can go to the hidden room "__CitadelSMTPSpoolout__" and your queue will be in there.
If you find delivery instructions that reference that message, just delete them.

Hi what are you meaning?you wrote lol 

Who are you and why write this things?

Hello,

After the upgrade I am playing around with the server a bit to harden and improve reputation. I noticed that TLS does not seem to be advertized. Tested with the local server:

~$ telnet localhost 25
Trying ::1...
Trying 127.0.0.1...
Connected to localhost.
Escape character is '^]'.
220 [my host] ESMTP Citadel server ready.
EHLO [some alias]
250-Hello [some alias] (localhost [127.0.0.1])
250-HELP
250-SIZE 10485760
250-AUTH LOGIN PLAIN
250-AUTH=LOGIN PLAIN
250 8BITMIME

The server has certificates installed and uses those when connecting via IMAP. There are also a lot of TLS messages in syslog. But in the response above I miss repsonse line 250-STARTTLS indicating TLS support for the MTA. STARTTLS command is accepted (220) though so in theory would work but as it's not advertized might not be used when remote server is waiting for that capability response.

Am I missing something?

Kind regards,

P.

Hello dear admin and support I hope you are in good mood today actually I found a important bug in citadel mail server and it could endanger sensitive data in local and global citadels I never ill-use that bug and it's information as you now no pain no gain I decided to inform you that report by email but can I expect prize ?(I am bug hunter and receive prize in trust wallet) your helping and cooperation can encourage me to trying find more bugs we can make better internet environment by the way I don't want too much money after observing and confirmation you can evaluate my request

We do not negotiate with extortionists.  If you believe you have found a bug in the software please post it here or private mail to me if you feel it is sensitive information.

Is there a way to remove the message number via command line client?

That might actually be a better option.  Once logged in to the command line client, type ".Goto __CtdlSMTPSpoolout__"

Then hit "F" to read-Forward

You'll see two types of messages there.  Half ought to be "delivery instructions" and the other half should be the messages themselves.  If you find one without the other it is a stray message and you can hit "D" to delete it.

BTW, is there a documentation on error and return codes somewhere? I noticed that 'systemctl stop citadel' does not exit clean but with retun code 15. Not sure what this means.

That's a different story.  But there's a way to find out.  Stop the Citadel server with systemctl, then start it in the foreground with:

cd /usr/local/citadel

./citserver -x9

Run it for a little while, then stop it with ctrl-c or by shutting it down from the user interface.  Then post the last screenful of log messages here, so we can see why you got that exit code.

After the upgrade I am playing around with the server a bit to harden and improve reputation. I noticed that TLS does not seem to be advertized. Tested with the local server:

The server has certificates installed and uses those when connecting via IMAP. There are also a lot of TLS messages in syslog. But in the response above I miss repsonse line 250-STARTTLS indicating TLS support for the MTA. STARTTLS command is accepted (220) though so in theory would work but as it's not advertized might not be used when remote server is waiting for that capability response.

Am I missing something?

You're not missing anything; that is the normal behavior.

The reason we do not advertise STARTTLS is because too many sites start up with a self-signed certificate, and we've found that those sites will not receive mail from many popular sources if STARTTLS is advertised but the certificate is invalid from the sender's perspective.

If this is something you need, we could add an undocumented server configuration item to switch it on manually.  Would that help?

Hello! I provide Citadel hosting starting at $7 and would like to know if I can get it listed it here. My website is at https://citadelhosting.retro-os.live/. If any questions I'm listening at fstltna@yahoo.com :)

cd /usr/local/citadel

./citserver -x9

Run it for a little while, then stop it with ctrl-c or by shutting it down from the user interface.  Then post the last screenful of log messages here, so we can see why you got that exit code.

citserver[251519]: Additional memory allocated since startup: 0 bytes
citserver[251519]: listdeliver: delivery interval not yet reached; last run was 2m2s ago
citserver[251519]: smtpclient: start full queue run , last_queue_job_processed=0 , last_queue_job_submitted=0
citserver[251519]: smtpclient: 0 messages to be processed
citserver[251519]: smtpclient: end full queue run , last_queue_job_processed=0 , last_queue_job_submitted=0
citserver[251519]: rssclient: polling interval not yet reached; last run was 2m2s ago
citserver[251519]: bdb: -- checkpoint --
citserver[251519]: Additional memory allocated since startup: 0 bytes
^Ccitserver[251519]: sysdep: caught signal 2
citserver[251519]: citserver: closing databases
citserver[251519]: bdb: performing final checkpoint
citserver[251519]: bdb: flushing the database logs
citserver[251519]: bdb: closing databases
citserver[251519]: bdb: closing database 00
citserver[251519]: bdb: closing database 01
citserver[251519]: bdb: closing database 02
citserver[251519]: bdb: closing database 03
citserver[251519]: bdb: closing database 04
citserver[251519]: bdb: closing database 05
citserver[251519]: bdb: closing database 06
citserver[251519]: bdb: closing database 07
citserver[251519]: bdb: closing database 08
citserver[251519]: bdb: closing database 09
citserver[251519]: bdb: closing database 0a
citserver[251519]: bdb: closing database 0b
citserver[251519]: bdb: closing database 0c
citserver[251519]: bdb: closing database 0d
citserver[251519]: bdb: closing environment
citserver[251519]: bdb: shutdown completed
citserver[251519]: citserver: Exiting with status 2
citserver[251519]: main: removing lockfile

Looks like this is creatign a different result than when run under systemd.

That might actually be a better option.  Once logged in to the command line client, type ".Goto __CtdlSMTPSpoolout__"

Then hit "F" to read-Forward

Indeed I found there two message that were more than a year old and deleted them. This greatly reduced the amount of noise in the log. But I still have msgbase errors for two messages. One comes up around every half an hour, the other has longer cycles.

Jan 18 20:38:41 [host] citserver[251656]: citserver[251656]: msgbase: message #93647 was not found
Jan 18 20:38:41 [host] citserver[251656]: msgbase: message #93647 was not found
Jan 18 20:38:41 [host] citserver[251656]: citserver[251656]: msgbase: message #93647 was not found
Jan 18 20:38:41 [host] citserver[251656]: citserver[251656]: msgbase: message #93647 was not found
Jan 18 20:38:41 [host] citserver[251656]: citserver[251656]: msgbase: error; invalid preloaded message for output
Jan 18 20:38:41 [host] citserver[251656]: msgbase: message #93647 was not found
Jan 18 20:38:41 [host] citserver[251656]: msgbase: message #93647 was not found
Jan 18 20:38:41 [host] citserver[251656]: msgbase: error; invalid preloaded message for output

Is there a way from the command line to identify a specific message number?

KR,

P.

If this is something you need, we could add an undocumented server configuration item to switch it on manually.  Would that help?

That would be great. I mainly use Citadel as mail and collaboration platform and this would be greatly improve reputation interacting with other public MTAs. Not mega urgent though probably something for a next release.

Many thanks for considering!

P.