If you know the actual name of the room (such as "8644012017.Mail") you can
use the private room commands to invite another user to it, but there's no
way to prevent the owner of the room from seeing that someone else has access
to the room.

Presumably related to the housekeeping loop issue, there seems to be a general problem with my Citadel system where idle sessions are not cleared up and just stay alive indefinitely. This is causing the CPU usage to max out and eventually bring down the entire box (not just citserver).

I noticed that the 'Maximum concurrent sessions' setting was unlimited (as far as I know it has always been set to 0), so I changed it to 20. However, Citadel soon hits 20 sessions and stops accepting SMTP mail or allowing IMAP logons or webcit logins until restarted manually.

syslog contains 'context: scheduled 20 idle sessions for termination'

The box it is running on is:
Operating system: Debian Linux 10 (32 bit)
Processor: Intel(R) Atom(TM) CPU D525 @ 1.80GHz, 4 cores
RAM: 2GB RAM

Is anyone else facing this issue or am I just lucky?

The problem still exists, I have  with a new installation of citadel 929 on
raspberrypi 3 with a fresh raspbian - based on debian buster.

I could sponsor a new raspberrypi to citadel.org to test it.

Fri May 08 2020 11:10:25 EDT from IGnatius T Foobar @ Uncensored Subject: Re: Corrupted attachments

I have a really old Pi and will see what I can do to test. It might not work at all, but who knows. What is the size of the attachment you tried to send?

I have Citadel installed on two networks sandboxed (no internet) and also on a network with internet. For some reason Citadel takes too long to load on the sandboxed network but no problems on the network with internet access. What could be the cause?

DNS resolution most likely, If you have no DNS service at all, a lot of services are going to be painfully slow to respond.

Try adding some local dns entries in /etc/hosts or router, or whatever is pretending to be your dns.. (that will fix slow ssh logins also.

192.168.0.5 mail mail.local.lan muh.mailserver.lan  citadel  # whatever

Thu Aug 20 2020 11:56:44 AM EDT from josel @ Uncensored Subject: Citadel on standalone network

I have Citadel installed on two networks sandboxed (no internet) and also on a network with internet. For some reason Citadel takes too long to load on the sandboxed network but no problems on the network with internet access. What could be the cause?

Forgot to mention that each network does have it's own DNS server.

When doing an nslookup by domain or IP I get an immediate response, same with curl or did commands. The latency is when I use the browser to get to Citadel mail.

Is the browser on the citadel box, or a different one?   Is there any delay connected locally or remotely with the text client? Try both.

Either case, I'd have a terminal open with top or htop, maybe another with tail -f syslog .. then hit webcit iwith the browser and watch what happens.   Get as much info together as you can.. Isolate the problem to a realm. 

Modern browsers do weird things. also, "extensions" like "ghostery" and "adblock" and "https everywhere" .  Slow to me says "resolution timeout" ..  not citserver.  which is always "really fast".

I notice I can't even hit a local server with a browser anymore until I literally type in "HTTP:\\mylocal.server", the full url, with scheme, at least ONE TIME.  Because most browsers want to assume it's public site, try to decide if i's a query, try to do a prefectch. ask you if you want to order from amazoom, send passwords to grooogle, or any other of the things that they like to do, so try it with something like lynx and see if you get the same result.  

That's how I would proceed.  Then compare your findings to your public box. It shouldn't be too hard to isolate.  Also, "right click" "inspect element"  "Network" and "reload" Check the actual request thats going out, and see whats its doing in the waterfall..

Might not be citadel at all. If it turns out to be that, I'd take a close look at your disk media.  That is, if dns does not turn out to be the issue after all.  I'd still make sure there are some entries in hosts on the server box EVEN IF YOU DON'T THINK YOU NEED THEM.

 A lot of things depend on that even with fake names and spoofed dns servers.  The key to understanding will be found by "ssh".  if it's slow to get to by ssh, you need a hosts entry for your loopback interface to "server.name" on the server box.

Thu Aug 20 2020 01:13:08 PM EDT from josel @ Uncensored Subject: Re: Citadel on standalone network

When doing an nslookup by domain or IP I get an immediate response, same with curl or did commands. The latency is when I use the browser to get to Citadel mail.

Browser is on a different box. I have also done what you suggested regarding right click and inspect. What I've seen is that it takes long long some .gif files.

Sorry, I meant that it takes long loading .gif files.

AHA!  I KNEW is was those bloody gifs! .. hahaha, thanks man for the gut laugh of the day...

Those tiny little image files should not be causing you any problems..  What is top showing when you hit the page? Are you tailing syslog? 

ssh mail.server [log in]

sudo su

tail -f /var/log/syslog

then..

ssh mail.server [log in]

top

THEN load the page.

:)

Thu Aug 20 2020 03:52:05 PM EDT from josel @ Uncensored Subject: Re: Citadel on standalone network

Browser is on a different box. I have also done what you suggested regarding right click and inspect. What I've seen is that it takes long long some .gif files.

All,

Did this ever get resolved?  I just tried this on my Citadel Server (Raspberry PI 3B+) and my text file produced:

TEXT
TEXT
TEXT
TEXT
TEXT
1234
1234
1234
1234
123ÿó@Ð

the Original:

TEXT
TEXT
TEXT
TEXT
TEXT
1234
1234
1234
1234
1234

So it appears to be something with appending to the file or? 

Any ideas?

Thanks!

Sat May 09 2020 23:13:54 EDT from warbaby @ Uncensored Subject: Re: Corrupted attachments

You might try a few more text uploads with various strings and see if there is any consistency with the char length.  

You might try creating some new files with a text editor and see how they look after a reboot.

I predict it is going to be failing Chink SD card writes, rather than a problem with Citadel. 

Or at least, replacing the Chink SD card would be my first approach, since that's a more widely known problem. 

Who knows all the various ways they fail? It might not be a total failure, just not writing properly.

Sometimes you can edit and save a file, it looks okay if you cat it out, but shows garbage after rebooting.

 

Sat May 09 2020 06:58:25 PM EDT from MAS3 @ Uncensored Subject: Re: Corrupted attachments

Thanks for the suggestion.

I created a small file, containing this text:

 

This is a text file created by the standard windows text editor.
It is not a large file, but i'm sure it will be corrupted once it's sent.

So let's see what will happen to it.

 

And this was what was received in the attachment (i'm not sure it will show exactly as i received due to the corruption)
Hmm seems there's more to it, as the font size has changed after copy / paste, and indentation was added:

 

This is a text file created by the standard windowsüÑ•áЁ•‘¥Ñ½È¸4)%Ё¥Ì¹½Ð„±…ɝ”™¥±”°‰ÕЁ¤´ÍÕÉ—ò—Bv–ÆÂ&R6÷''WFVBöæ6R—Bw26VçBàРХ6òÆWBw2?ÙYHÚ]Ú[\[ˆÈ]ƒBþ¾

 

it has something to do with character encoding I think.

My editor gedit says, that the original file I sent ist UTF-8 encoded, the file I downloaded with webcit is ISO 8859-15 encoded. See the files attached.

Does this only occur for German installations ?

Mon Aug 24 2020 10:22:18 EDT from phaake37 @ Uncensored Subject: Re: Corrupted attachments

All,

 

Did this ever get resolved?  I just tried this on my Citadel Server (Raspberry PI 3B+) and my text file produced:

 

TEXT
TEXT
TEXT
TEXT
TEXT
1234
1234
1234
1234
123ÿó@Ð

 

the Original:

TEXT
TEXT
TEXT
TEXT
TEXT
1234
1234
1234
1234
1234

So it appears to be something with appending to the file or? 

 

Any ideas?

Thanks!

 

 

Sat May 09 2020 23:13:54 EDT from warbaby @ Uncensored Subject: Re: Corrupted attachments

You might try a few more text uploads with various strings and see if there is any consistency with the char length.  

You might try creating some new files with a text editor and see how they look after a reboot.

I predict it is going to be failing Chink SD card writes, rather than a problem with Citadel. 

Or at least, replacing the Chink SD card would be my first approach, since that's a more widely known problem. 

Who knows all the various ways they fail? It might not be a total failure, just not writing properly.

Sometimes you can edit and save a file, it looks okay if you cat it out, but shows garbage after rebooting.

 

Sat May 09 2020 06:58:25 PM EDT from MAS3 @ Uncensored Subject: Re: Corrupted attachments

Thanks for the suggestion.

I created a small file, containing this text:

 

This is a text file created by the standard windows text editor.
It is not a large file, but i'm sure it will be corrupted once it's sent.

So let's see what will happen to it.

 

And this was what was received in the attachment (i'm not sure it will show exactly as i received due to the corruption)
Hmm seems there's more to it, as the font size has changed after copy / paste, and indentation was added:

 

This is a text file created by the standard windowsüÑ•áЁ•‘¥Ñ½È¸4)%Ё¥Ì¹½Ð„±…ɝ”™¥±”°‰ÕЁ¤´ÍÕÉ—ò—Bv–ÆÂ&R6÷''WFVBöæ6R—Bw26VçBàРХ6òÆWBw2?ÙYHÚ]Ú[\[ˆÈ]ƒBþ¾

 

 

blub2.txt (text/plain, 95 bytes) [View| Download]

hallo

hallo
hallop
hhh

mnnb
öö
ioio

äääüü

€€€€

blub2webcit.txt (text/plain, 99 bytes) [View| Download]

hallo

hallo
hallop
hhh

mnnb
öö
ioio

äääüü

���*�(*�(*�(*����

I am trying to install Citadel using EasyInstall on a Ubuntu VM on Windows 10.

Here's what I am receiving... Can anyone provide any suggestions?

MyServer:~$ sudo -s

root@MyServer:/home/andrew# curl http://easyinstall.citadel.org/install | bash

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current

                                 Dload  Upload   Total   Spent    Left  Speed

100   185  100   185    0     0    642      0 --:--:-- --:--:-- --:--:--   642

bash: line 1: html: No such file or directory

bash: line 2: syntax error near unexpected token `<'

'ash: line 2: `<head><title>301 Moved Permanently</title></head>

Hi Andrew! 

If you look closely you can see clearly you got html back instead of a script.

Try:

wget http://easyinstall.citadel.org/install --output-document=easy-install.sh

or use the capital O switch

wget http://easyinstall.citadel.org/install --O easy-install.sh

then you can

chmod +x easy-install.sh

./easy-install.sh

Wed Sep 02 2020 10:00:03 AM EDT from vogelap @ Uncensored Subject: Trying to install on Ubuntu VM in Windows 10

I am trying to install Citadel using EasyInstall on a Ubuntu VM on Windows 10.

Here's what I am receiving... Can anyone provide any suggestions?

 

MyServer:~$ sudo -s

root@MyServer:/home/andrew# curl http://easyinstall.citadel.org/install | bash

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current

                                 Dload  Upload   Total   Spent    Left  Speed

100   185  100   185    0     0    642      0 --:--:-- --:--:-- --:--:--   642

bash: line 1: html: No such file or directory

bash: line 2: syntax error near unexpected token `<'

'ash: line 2: `<head><title>301 Moved Permanently</title></head>

 

Woops, that should only be one - on the dash O example. 

wget ... -O easy-install.sh

Wed Sep 02 2020 10:08:18 AM EDT from warbaby @ Uncensored Subject: Re: Trying to install on Ubuntu VM in Windows 10

Hi Andrew! 

If you look closely you can see clearly you got html back instead of a script.

Try:

wget http://easyinstall.citadel.org/install --output-document=easy-install.sh

or use the capital O switch

wget http://easyinstall.citadel.org/install --O easy-install.sh

then you can

chmod +x easy-install.sh

./easy-install.sh

 

Wed Sep 02 2020 10:00:03 AM EDT from vogelap @ Uncensored Subject: Trying to install on Ubuntu VM in Windows 10

I am trying to install Citadel using EasyInstall on a Ubuntu VM on Windows 10.

Here's what I am receiving... Can anyone provide any suggestions?

 

MyServer:~$ sudo -s

root@MyServer:/home/andrew# curl http://easyinstall.citadel.org/install | bash

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current

                                 Dload  Upload   Total   Spent    Left  Speed

100   185  100   185    0     0    642      0 --:--:-- --:--:-- --:--:--   642

bash: line 1: html: No such file or directory

bash: line 2: syntax error near unexpected token `<'

'ash: line 2: `<head><title>301 Moved Permanently</title></head>

 

 

How would I automatically (or in a scripted or other way) automatically add/collect all sender's email addresses from the email header and then add it to the Address Book, Mailing List, or database file?

Thank you for your support.

Hello, I've been happily using Citadel for my family's mail server for years, and have not had but a very few problems, and every time simply restoring from a backup has cured it.

This time, however, whatever happened seems to have persisted through most of my backups while the server kept running, so almost all my backups are corrupted.

I _do_ have a good backup from mid-August, but I'd much rather not have to export all my family member's IMAP client data (or likely lose all that email)

Here are the errors generated after restoring the last 5 days worth of citadel/data:

root@mail:~# cat /var/log/mail.err | cut -d':' -f4- | sort | uniq -c | sort -n
   4929  DB: BDB2506 file cdb.00 has LSN 14765/1303875, past end of log at 1/28
   4970  DB: BDB2506 file cdb.00 has LSN 14795/4275094, past end of log at 1/28
   9053  DB: BDB2506 file cdb.00 has LSN 14826/1401646, past end of log at 1/28
   9520  DB: BDB2506 file cdb.00 has LSN 14889/4245711, past end of log at 1/28
  12086  DB: BDB2506 file cdb.00 has LSN 14856/9334939, past end of log at 1/28

As you can see from the uniq -c, it simply spams that single message in my mail.err log until I shut down the service.

I can't access citserver in any way that I know of while its running...

Are there any tools to save newer versions of my backups, or is it hopeless?

Thanks!

    -Brian

Hello, thanks for the recovery advice.

I've tried against all existing backups I have (7 days worth) but they have all failed to produce a citadel instance that will start.

I've run the database_cleanup.sh script, which proceeded normally with no errors or otherwise unusual looking output.

Running db_recover now returns the following:

root@mail:~# db_recover -cvvvvv
BDB2526 Finding last valid log LSN: file: 1 offset 28

Here's the output when I try to run citserver:

root@mail:~# /usr/sbin/citserver -x2 -lmail
citserver[2306]: 
citserver[2306]: 
citserver[2306]: *** Citadel server engine ***
citserver[2306]: Version 902 (build 1a4ec92) ***
citserver[2306]: Copyright (C) 1987-2016 by the Citadel development team.
citserver[2306]: This program is distributed under the terms of the GNU General Public License.
citserver[2306]: 
citserver[2306]: Called as: /usr/sbin/citserver
citserver[2306]: libcitadel(unnumbered)
citserver[2306]: master_startup() started
citserver[2306]: Checking directory access
citserver[2306]: Opening databases
citserver[2306]: bdb(): open_databases() starting
citserver[2306]: Compiled db: Berkeley DB 5.3.28: (September  9, 2013)
citserver[2306]:   Linked db: Berkeley DB 5.3.28: (September  9, 2013)
citserver[2306]: Linked zlib: 1.2.11
citserver[2306]: bdb(): Setting up DB environment
citserver[2306]: dbenv->open(dbenv, /var/lib/citadel/data/, 75043, 0)
citserver[2306]: DB: BDB2526 Finding last valid log LSN: file: 1 offset 67923
citserver[2306]: DB: BDB1514 Recovery starting from [1][62827]
citserver[2306]: DB: BDB1518 Recovery complete at Tue Sep  8 17:18:03 2020
citserver[2306]: DB: BDB1519 Maximum transaction ID 80000021 recovery checkpoint [1][69085]
citserver[2306]: Starting up DB
citserver[2306]: Initializing configuration system
citserver[2306]: configuration setting c_default_cal_zone is empty, but must not - check your config!
citserver[2306]: Checking floor reference counts

Message from syslogd@mail at Sep  8 17:18:03 ...
 citserver[2306]: configuration setting c_default_cal_zone is empty, but must not - check your config!
citserver[2306]: Floor 0: 291 rooms
citserver[2306]: Floor 1: 0 rooms
citserver[2306]: Floor 2: 0 rooms
citserver[2306]: Floor 3: 0 rooms
citserver[2306]: Floor 4: 0 rooms
citserver[2306]: Floor 5: 0 rooms
citserver[2306]: Floor 6: 0 rooms
citserver[2306]: Floor 7: 0 rooms
citserver[2306]: Floor 8: 0 rooms
citserver[2306]: Floor 9: 0 rooms
citserver[2306]: Floor 10: 0 rooms
citserver[2306]: Floor 11: 0 rooms
citserver[2306]: Floor 12: 0 rooms
citserver[2306]: Floor 13: 0 rooms
citserver[2306]: Floor 14: 0 rooms
citserver[2306]: Floor 15: 0 rooms
citserver[2306]: Creating base rooms (if necessary)
citserver[2306]: Seeding the pseudo-random number generator...
citserver[2306]: master_startup() finished
citserver[2306]: Sanity checking the recorded highest message, user, and room numbers
citserver[2306]: bdb(): cursor still in progress on cdb 02: attempt to write during r/o cursor
citserver[2306]: citserver is stopping in order to prevent data loss. uid=0 gid=0 euid=0 egid=0

Any advice for other things to try would be much appreciated, as I found at least one user that only used webcit and has no local imap copy of their data!

Thanks again!

   -Brian