Language:
switch to room list switch to menu My folders
Go to page: [1] 2 3 4 5 ... Last
↑↑↑ Old messages ↑↑↑            ↓↓↓ New messages ↓↓↓
[#] Wed May 07 2014 03:41:23 EDT from rolfijn @ Uncensored

Subject: Re: How to edit a draft ?

[Reply] [ReplyQuoted] [Headers] [Print]

Yes! Exellent question. Is there a way? I've also not been able to find it.

Sun Dec 29 2013 14:49:24 EST from mingodad @ Uncensored Subject: How to edit a draft ?

I'm just starting with citadel and I like the concept but I can not understand how we can edit a draft document, I can only see options to reply, forward not to edit the draft itself.

How it's supposed to work ?

Thanks in advance for any help !



 



[#] Fri May 09 2014 14:41:12 EDT from lee @ Uncensored

Subject: Admin problems

[Reply] [ReplyQuoted] [Headers] [Print]

Hi, I've just installed and don't seem to have global admin privileges. It asked me when installing to enter a password for admin but when I tried to login after installation it was as if the admin user didn't exist. I have to create an account called admin but I can't see any global options in the advanced settings.



[#] Fri May 09 2014 15:20:37 EDT from Freakdog @ Dog Pound BBS II

[Reply] [ReplyQuoted] [Headers] [Print]

lee: It also asked you for the username to which admin privileges were supposed to be granted.

You need to create an account in your citadel with that username.

[#] Sat May 10 2014 15:58:08 EDT from "Jacek Zaleski" <sirjacek@gmail.com> to Citadel Support <citadel_support@citadel.org>

Subject: Anyone have this same trouble with stoping server

[Reply] [ReplyQuoted] [Headers] [Print]

 DB: /var/lib/citadel/data/log.0000012887: log file unreadable: Too many open files
May  9 23:41:43 atlant citserver[32270]: DB: PANIC: Too many open files
May  9 23:41:43 atlant citserver[32270]: bdb(): PANIC: Too many open files
May  9 23:42:32 atlant citserver[32270]: Network full processing in 1 seconds.
May  9 23:42:32 atlant citserver[32270]: DB: PANIC: fatal region error detected; run recovery
May  9 23:42:32 atlant citserver[32270]: bdb(): PANIC: DB_RUNRECOVERY: Fatal error, run database recovery
May  9 23:42:32 atlant citserver[32270]: localcursor: DB_RUNRECOVERY: Fatal error, run database recovery
May  9 23:42:32 atlant citserver[32270]: citserver is stopping in order to prevent data loss. uid=106 gid=110 euid=106 egid=110

[
GG    :  3857388

Before printing this document , think about the environment


]


[#] Sun May 11 2014 09:14:13 EDT from dothebart @ Uncensored

Subject: Re: Anyone have this same trouble with stoping server

[Reply] [ReplyQuoted] [Headers] [Print]

 

Sat May 10 2014 15:58:08 EDT from "Jacek Zaleski" <sirjacek@gmail.com> Subject: Anyone have this same trouble with stoping server
 DB: /var/lib/citadel/data/log.0000012887: log file unreadable: Too many open files
May  9 23:41:43 atlant citserver[32270]: DB: PANIC: Too many open files
May  9 23:41:43 atlant citserver[32270]: bdb(): PANIC: Too many open files
May  9 23:42:32 atlant citserver[32270]: Network full processing in 1 seconds.
May  9 23:42:32 atlant citserver[32270]: DB: PANIC: fatal region error detected; run recovery
May  9 23:42:32 atlant citserver[32270]: bdb(): PANIC: DB_RUNRECOVERY: Fatal error, run database recovery
May  9 23:42:32 atlant citserver[32270]: localcursor: DB_RUNRECOVERY: Fatal error, run database recovery
May  9 23:42:32 atlant citserver[32270]: citserver is stopping in order to prevent data loss. uid=106 gid=110 euid=106 egid=110
 
[
GG    :  3857388

Hm, no not yet seen...

can you have a look at lsof which are those files?

potentially it could be a spam attack; adding a rate limit for conections in iptables.

which citadel version are you running?



[#] Sun May 11 2014 21:19:44 EDT from Freakdog @ Dog Pound BBS II

[Reply] [ReplyQuoted] [Headers] [Print]

I've been having more than a few of those DB crashes, lately, running 8.24. I figured it had something to do with my having restored my server and then the VM in which the Citadel ran.

[#] Sun May 11 2014 21:32:49 EDT from Freakdog @ Dog Pound BBS II

[Reply] [ReplyQuoted] [Headers] [Print]

Anyhow...having read the suggestion in question, and seeing the same error in my log (log file unreadable, too many files open), I've added a rate limit on port 25 on my BBS VM...we'll see how that works out for me.

[#] Mon May 12 2014 01:31:28 EDT from TheOneLaw @ Uncensored

Subject: Global Address Book redundant remote Server citadel networking

[Reply] [ReplyQuoted] [Headers] [Print]

We did this back in 2010 and it worked very well, version 7.86 (?)

Until just this past month I had not revisited this issue,

 but it would be convenient if we could, or had another way to do it.

 

Now I have been trying to do it again this month and the main server (v8.24) crashes on a 1-minute loop,

 so I had to disable the replication service, which instantly fixed that  particular problem.

I assume I have other issues with the setup, and in light of the deprecation will not worry to much about it anyway.

 

Maybe the old Qmail 'POP Toaster' is what I need to try to set up,

 but fundamentally what we want to achieve is having our guys in the jungle able to access their local server

 complete with all their stored mail even when said server has its link to home broken (as frequently happens)

They will have operations and personnel interconnected over several hundred square kilometers,

 with internet accounts which will need to have outgoing mail held for the moments when the server links are up.

 

A remotely distributed email server with replicated accounts and all their emails stored redundantly.

 

Cheers, anyway,

whatever you do I am sure it will be a step forward.

 


Sun Mar 16 2014 09:44:54 EDT from "ben-Nabiy Derush" <bennabiy@msrfarm.com> Subject: Re: [Citadel Support] (no subject)
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

We currently are using citadel to citadel communication to remove
outside email and use citadel for communication between two domains
only. Would this be interrupted?

We have internet mail off, but use the smtp features for local users
to mail each other, but not outside domains, with the exception of the
other citadel domain. We share a Global Address book, and for all
intensive purposes have the two machines mirroring each other, except
for local mail stores ( they both can function independently as well).

Is this going to be broken?

On 01/03/2014 06:35 PM, IGnatius T Foobar wrote:

We're looking to do a fairly significant data model change, so I
wanted to check to make sure no one is using a feature which we
would like to deprecate:


Are there any known cases where someone has an Internet email
domain spread out across multiple Citadel servers, by sharing the
Global Address Book room and making every Citadel an SMTP server?

This was intended to build distributed systems, but I don't know of
anyone actually using this feature. We would like to remove it.
In fact, we would like to remove Citadel-to-Citadel networking
entirely in its current form, and replace it with NNTP as the
primary means of sharing rooms between Citadel sites.

If anyone actually has an installation where Global Address Book is
shared and being used for a distributed email domain, please let me
know immediately so we can find an ongoing solution for you. In
the meantime please consider this feature deprecated, and it will
be removed in Citadel 9.


- --
??????? ?????? ??? ?????
ben-Nabiy Derush of Yowceph
The Community Conference Center
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.21 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBAgAGBQJTJarWAAoJEH5nVBNX4CTJ33QH/i1g6maILBVc+QXx2gyWpKKO
ofI0S6gY0l36Xyj94x1A9hqMnYaQ2Kh6IeCoA77I3m9RjaMLh2UJi8uDXdMVaK2l
hfT3LC+T2ooyHlj0JYYdSktGpsF6B8/8yFcM/xj7bh8gZx+5RkQ/vqQRPLL9Gu+b
xZl5luwNt1R70Q8LKo7EbCPl/2GgjicfG6UqETVAHcYoQL+1cpYwlKvQyF3KjwNy
kqaTYjm7S5J29puv7401UHpddVvDnnQVXykV72iFbvsmCpHpHg4H9aioPF4rVhdK
jqIV0iCKe+qWJ2rfH+U9xC6X9QWS38rknZxW8y4tYFWQdmalMWtsChN7G5b2Cxg=
=nNke
-----END PGP SIGNATURE-----


-- 
TheOneLaw



[#] Mon May 12 2014 06:27:28 EDT from "Jacek Zaleski" <sirjacek@gmail.com> to Citadel Support <citadel_support@citadel.org>

Subject: Re: [Citadel Support]

[Reply] [ReplyQuoted] [Headers] [Print]

Version cit is Citadel server engine v8.14 (build 53481be) 
~# lsof -a /var/lib/citadel/data/*
COMMAND    PID    USER   FD   TYPE DEVICE  SIZE/OFF   NODE NAME
citserver 7346 citadel    4uW  REG    9,1        56 919782 /var/lib/citadel/data/citadel.control
citserver 7346 citadel    5u   REG    9,1 144261120 919785 /var/lib/citadel/data/cdb.00
citserver 7346 citadel    6u   REG    9,1      8192 919786 /var/lib/citadel/data/cdb.01
citserver 7346 citadel    7u   REG    9,1     32768 919787 /var/lib/citadel/data/cdb.02
citserver 7346 citadel    8u   REG    9,1     16384 919788 /var/lib/citadel/data/cdb.03
citserver 7346 citadel    9u   REG    9,1   1048576 919789 /var/lib/citadel/data/cdb.04
citserver 7346 citadel   10u   REG    9,1     20480 919790 /var/lib/citadel/data/cdb.05
citserver 7346 citadel   11u   REG    9,1      8192 919791 /var/lib/citadel/data/cdb.06
citserver 7346 citadel   12u   REG    9,1      8192 919792 /var/lib/citadel/data/cdb.07
citserver 7346 citadel   13u   REG    9,1 439566336 919793 /var/lib/citadel/data/cdb.08
citserver 7346 citadel   14u   REG    9,1  61435904 919795 /var/lib/citadel/data/cdb.09
citserver 7346 citadel   15u   REG    9,1      8192 919796 /var/lib/citadel/data/cdb.0a
citserver 7346 citadel   16u   REG    9,1      8192 917339 /var/lib/citadel/data/cdb.0b
citserver 7346 citadel   17u   REG    9,1      8192 919798 /var/lib/citadel/data/cdb.0c
citserver 7346 citadel   18u   REG    9,1  10485760 928212 /var/lib/citadel/data/log.0000012903


[
GG    :  3857388

Before printing this document , think about the environment


]


On Sun, May 11, 2014 at 3:15 PM, Citadel Support <citadel_support@citadel.org> wrote:
 ------------------------------------------------------------------------------
From: dothebart <dothebart@uncensored.citadel.org>
Subject: Re: Anyone have this same trouble with stoping server

 
> Sat May 10 2014 15:58:08 EDT from "Jacek Zaleski" <sirjacek@gmail.com>
>Subject: Anyone have this same trouble with stoping server
>
>    DB: /var/lib/citadel/data/log.0000012887: log file unreadable: Too many
>open files
>
> May  9 23:41:43 atlant citserver[32270]: DB: PANIC: Too many open files
>
> May  9 23:41:43 atlant citserver[32270]: bdb(): PANIC: Too many open files
>
> May  9 23:42:32 atlant citserver[32270]: Network full processing in 1
>seconds.
>
> May  9 23:42:32 atlant citserver[32270]: DB: PANIC: fatal region error
>detected; run recovery
>
> May  9 23:42:32 atlant citserver[32270]: bdb(): PANIC: DB_RUNRECOVERY:
>Fatal error, run database recovery
>
> May  9 23:42:32 atlant citserver[32270]: localcursor: DB_RUNRECOVERY:
>Fatal error, run database recovery
>
> May  9 23:42:32 atlant citserver[32270]: citserver is stopping in order to
>prevent data loss. uid=106 gid=110 euid=106 egid=110
>
>  
>
> [
>GG    :  3857388
>
>
>
>
>
>



Hm, no not yet seen...

can you have a look at lsof which are those files?

potentially it could be a spam attack; adding a rate limit for conections in
iptables.

which citadel version are you running?
 ------------------------------------------------------------------------------
You are subscribed to the 'Citadel Support' list.
To post to the list: citadel_support@citadel.org



[#] Mon May 12 2014 15:32:55 EDT from herrinternet @ Uncensored

Subject: server-to-server-sync project question

[Reply] [ReplyQuoted] [Headers] [Print]

hello everyone, 

i am a proud user of the citadel groupware project for a year now and i am amazed by its robust and light-footed approach to email and groupware functions in general. Now that me and some friends run citadel on a virtualized ubuntu machine the basement of a friend of mine, we would like to have local backups of our emails at home. 

I stumbled upon the synchronisation features of citadel and i am a bit confused.. Here is my/our setup: 

 

CitadelServer:

  • Enterprise Server Hardware
  • Public Static IP
  • LDAP-Server
  • Owncloud
  • Roundcube as Webmailer for Citadel

 

Here is what i would like to do, maybe you could give me a feedback if that is even possible:

I would like to sync the private rooms of the users (email, contacts, etc.) to specific citadel-servers within a vpn or ssh-tunnel through the web. Every User would have a local backup of their mails, boards etc. on the local citadel-server. 

Outgoing Email is also synced with the main-citadel server due its static ip-adress and good internet-connection. 

So my question consists of different questions:

  1. Is it possible to network citadel instances without opened wan-ports and a dynamic ip-adress (vpn/ssh-tunnel)?
  2. Is it possible to sync the private mail-rooms of the users to the networked citadel-instances?
  3. Is it possible to route email so that it is always sent with the public ip-adress of the "master" citadel-instance?
  4. Will the user-management of the "master" instance which is based on open-ldap scale or do the "slave" instances have to authenticate against the original ldap server?

Thank you very much for your time reading and maybe answering my question - and a big thank you to all the people involved in this great project - which provides simple and fun groupware-capability for a bunch of nerds who are fed up with the ususal enterprise owned cloud-based communication platforms.

Best Regards

Gregor aka. herr_internet

 

 



[#] Mon May 12 2014 17:53:21 EDT from dothebart @ Uncensored

Subject: Re: [Citadel Support]

[Reply] [ReplyQuoted] [Headers] [Print]

its not just berkeley DB which consumes file handles. rather grep for the PID of citserver than the files.

if many of them is internet connections, there you are.

Mon May 12 2014 06:27:28 EDT from "Jacek Zaleski" <sirjacek@gmail.com> Subject: Re: [Citadel Support]
Version cit is Citadel server engine v8.14 (build 53481be) 
~# lsof -a /var/lib/citadel/data/*
COMMAND    PID    USER   FD   TYPE DEVICE  SIZE/OFF   NODE NAME


[#] Mon May 12 2014 17:55:34 EDT from dothebart @ Uncensored

Subject: Re: Global Address Book redundant remote Server citadel networking

[Reply] [ReplyQuoted] [Headers] [Print]

please give a little more information.

http://citadel.org/doku.php/faq:mastering_your_os:gdb

Mon May 12 2014 01:31:28 EDT from TheOneLaw @ Uncensored Subject: Global Address Book redundant remote Server citadel networking

We did this back in 2010 and it worked very well, version 7.86 (?)

Until just this past month I had not revisited this issue,

 but it would be convenient if we could, or had another way to do it.

 

Now I have been trying to do it again this month and the main server (v8.24) crashes on a 1-minute loop,

 so I had to disable the replication service, which instantly fixed that  particular problem.

I assume I have other issues with the setup, and in light of the deprecation will not worry to much about it anyway.

 

Maybe the old Qmail 'POP Toaster' is what I need to try to set up,

 but fundamentally what we want to achieve is having our guys in the jungle able to access their local server

 complete with all their stored mail even when said server has its link to home broken (as frequently happens)

They will have operations and personnel interconnected over several hundred square kilometers,

 with internet accounts which will need to have outgoing mail held for the moments when the server links are up.

 

A remotely distributed email server with replicated accounts and all their emails stored redundantly.

 

Cheers, anyway,

whatever you do I am sure it will be a step forward.

 


Sun Mar 16 2014 09:44:54 EDT from "ben-Nabiy Derush" <bennabiy@msrfarm.com> Subject: Re: [Citadel Support] (no subject)
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

We currently are using citadel to citadel communication to remove
outside email and use citadel for communication between two domains
only. Would this be interrupted?

We have internet mail off, but use the smtp features for local users
to mail each other, but not outside domains, with the exception of the
other citadel domain. We share a Global Address book, and for all
intensive purposes have the two machines mirroring each other, except
for local mail stores ( they both can function independently as well).

Is this going to be broken?

On 01/03/2014 06:35 PM, IGnatius T Foobar wrote:

We're looking to do a fairly significant data model change, so I
wanted to check to make sure no one is using a feature which we
would like to deprecate:


Are there any known cases where someone has an Internet email
domain spread out across multiple Citadel servers, by sharing the
Global Address Book room and making every Citadel an SMTP server?

This was intended to build distributed systems, but I don't know of
anyone actually using this feature. We would like to remove it.
In fact, we would like to remove Citadel-to-Citadel networking
entirely in its current form, and replace it with NNTP as the
primary means of sharing rooms between Citadel sites.

If anyone actually has an installation where Global Address Book is
shared and being used for a distributed email domain, please let me
know immediately so we can find an ongoing solution for you. In
the meantime please consider this feature deprecated, and it will
be removed in Citadel 9.


- --
??????? ?????? ??? ?????
ben-Nabiy Derush of Yowceph
The Community Conference Center
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.21 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iQEcBAEBAgAGBQJTJarWAAoJEH5nVBNX4CTJ33QH/i1g6maILBVc+QXx2gyWpKKO
ofI0S6gY0l36Xyj94x1A9hqMnYaQ2Kh6IeCoA77I3m9RjaMLh2UJi8uDXdMVaK2l
hfT3LC+T2ooyHlj0JYYdSktGpsF6B8/8yFcM/xj7bh8gZx+5RkQ/vqQRPLL9Gu+b
xZl5luwNt1R70Q8LKo7EbCPl/2GgjicfG6UqETVAHcYoQL+1cpYwlKvQyF3KjwNy
kqaTYjm7S5J29puv7401UHpddVvDnnQVXykV72iFbvsmCpHpHg4H9aioPF4rVhdK
jqIV0iCKe+qWJ2rfH+U9xC6X9QWS38rknZxW8y4tYFWQdmalMWtsChN7G5b2Cxg=
=nNke
-----END PGP SIGNATURE-----

 

-- 
TheOneLaw



 



[#] Wed May 14 2014 09:26:07 EDT from Freakdog @ Dog Pound BBS II

Subject: DB: Panic: Too many open files.

[Reply] [ReplyQuoted] [Headers] [Print]

Citadel 8.24...this only cropped up after I had to do a full system restore on the host server and then the VM that runs the Citadel instance. Unfortunately, this seems to be happening daily.

May 14 03:07:17 bbs citserver[26719]: NW[comalies][2991]: polling
May 14 03:07:17 bbs citserver[26719]: NW[uncnsrd][2992]: polling
May 14 03:07:17 bbs citserver[26719]: NW[trada][2993]: polling
May 14 03:07:17 bbs citserver[26719]: serv_extnotify: processing notify queue
May 14 03:07:17 bbs citserver[26719]: serv_extnotify: queue run completed
May 14 03:07:17 bbs citserver[26719]: -- db checkpoint --
May 14 03:07:17 bbs citserver[26719]: not sending message again
May 14 03:07:17 bbs citserver[26719]: not sending message again
May 14 03:07:17 bbs citserver[26719]: not sending message again
May 14 03:07:18 bbs citserver[26719]: Context: [5429]SRV[CitNetworker] Session e
nded.
May 14 03:07:18 bbs citserver[26719]: Context: [5430]SRV[CitNetworker] Session e
nded.
May 14 03:07:18 bbs citserver[26719]: Context: [5431]SRV[CitNetworker] Session e
nded.
May 14 03:07:29 bbs citserver[26719]: DB: /usr/local/citadel/data/log.0000000290
: log file unreadable: Too many open files
May 14 03:07:29 bbs citserver[26719]: DB: PANIC: Too many open files
May 14 03:07:29 bbs citserver[26719]: bdb(): PANIC: Too many open files
May 14 03:08:18 bbs citserver[26719]: Network full processing in 412 seconds.
May 14 03:08:18 bbs citserver[26719]: DB: PANIC: fatal region error detected; run recovery
May 14 03:08:18 bbs citserver[26719]: bdb(): PANIC: DB_RUNRECOVERY: Fatal error, run database recovery
May 14 03:08:18 bbs citserver[26719]: localcursor: DB_RUNRECOVERY: Fatal error, run database recovery
May 14 03:08:18 bbs citserver[26719]: citserver is stopping in order to prevent data loss. uid=505 gid=505 euid=505 egid=505

I've already run the database_cleanup.sh script a number of times, (though not after this morning's event) to no avail.

Here's a listing of my data directory:

-rw------- 1 bbs root 185655296 May 14 09:00 cdb.00
-rw------- 1 bbs root      8192 May 14 09:00 cdb.01
-rw------- 1 bbs root     69632 May 14 09:00 cdb.02
-rw------- 1 bbs root     16384 May 14 03:09 cdb.03
-rw------- 1 bbs root   1089536 May 14 09:00 cdb.04
-rw------- 1 bbs root    126976 May 14 08:51 cdb.05
-rw------- 1 bbs root      8192 May  2 10:04 cdb.06
-rw------- 1 bbs root 443854848 May 14 09:09 cdb.07
-rw------- 1 bbs root  95764480 May 14 09:00 cdb.08
-rw------- 1 bbs root      8192 Apr 25 08:50 cdb.09
-rw------- 1 bbs root    184320 May 14 07:15 cdb.0a
-rw------- 1 bbs root      8192 May 14 03:09 cdb.0b
-rw------- 1 bbs root      8192 Apr 25 08:50 cdb.0c
-rw-rw---- 1 bbs bbs   10485760 May 14 09:09 log.0000000293

I wouldn't imagine that 177MB for cdb.00 or 423MB for cdb.07 should be too large for the DB engine...I'm sure that the DB files on Uncensored must be far larger.

I took the liberty, a couple of days ago, of increasing the nofiles ulimit for the "bbs" user from the default 1024 to 1536, though lsof only shows 417 "REG" entries.

Digging a little deeper, though, I'm seeing that 360 of those lsof hits are for one particular file in my /usr/local/citadel/network/spoolout directory (the spool file for the trada node, and that file is only 721.6K).

I've restarted the citadel and am watching the number of file handles (per lsof) grow, and the only network spool file is trada. Wondering if it's an issue with this particular spoolfile, I'm renaming the file and waiting to see if this crops up again.



[#] Wed May 14 2014 09:57:21 EDT from Jigar patel @ Uncensored

Subject: I can not receive mail? Installed Citadel

[Reply] [ReplyQuoted] [Headers] [Print]

I got Citadel installed and working and I have my user account set up, but for some reason it's not working 100%. I can send mail FROM my server, and receive it elsewhere, but I cannot send mail TO my server and have it show up anywhere.



[#] Wed May 14 2014 17:00:07 EDT from IGnatius T Foobar @ Uncensored

Subject: Re: I can not receive mail? Installed Citadel

[Reply] [ReplyQuoted] [Headers] [Print]

I got Citadel installed and working and I have my user account set
up, but for some reason it's not working 100%. I can send mail FROM
my server, and receive it elsewhere, but I cannot send mail TO my
server and have it show up anywhere.

If you telnet to port 25 on your server do you get an SMTP prompt that says "citadel" in the greeting?

If not, you have some other mail server running on your system and you need to shut it down.

[#] Thu May 15 2014 00:42:29 EDT from Jigar patel @ Uncensored

Subject: Re: I can not receive mail? Installed Citadel

[Reply] [ReplyQuoted] [Headers] [Print]

Is says

>telnet localhost 25

"ESMTP Citadel server ready."

from my pc

>telnet 212.71.235.233 25

Trying 212.71.235.233...
Connected to 212.71.235.233.
Escape character is '^]'.
220 cicor.hyperstatus.com ESMTP Citadel server ready.

But still same problem. help plz.



[#] Thu May 15 2014 11:15:33 EDT from dothebart @ Uncensored

Subject: Re: I can not receive mail? Installed Citadel

[Reply] [ReplyQuoted] [Headers] [Print]

[#] Thu May 15 2014 11:17:19 EDT from dothebart @ Uncensored

Subject: Re: DB: Panic: Too many open files.

[Reply] [ReplyQuoted] [Headers] [Print]

 

Wed May 14 2014 09:26:07 EDT from Freakdog @ Dog Pound BBS II Subject: DB: Panic: Too many open files.

Citadel 8.24...this only cropped up after I had to do a full system restore on the host server and then the VM that runs the Citadel instance. Unfortunately, this seems to be happening daily.

May 14 03:07:17 bbs citserver[26719]: NW[comalies][2991]: polling
May 14 03:07:17 bbs citserver[26719]: NW[uncnsrd][2992]: polling
May 14 03:07:17 bbs citserver[26719]: NW[trada][2993]: polling
May 14 03:07:17 bbs citserver[26719]: serv_extnotify: processing notify queue

I wouldn't imagine that 177MB for cdb.00 or 423MB for cdb.07 should be too large for the DB engine...I'm sure that the DB files on Uncensored must be far larger.

I took the liberty, a couple of days ago, of increasing the nofiles ulimit for the "bbs" user from the default 1024 to 1536, though lsof only shows 417 "REG" entries.

Digging a little deeper, though, I'm seeing that 360 of those lsof hits are for one particular file in my /usr/local/citadel/network/spoolout directory (the spool file for the trada node, and that file is only 721.6K).

I've restarted the citadel and am watching the number of file handles (per lsof) grow, and the only network spool file is trada. Wondering if it's an issue with this particular spoolfile, I'm renaming the file and waiting to see if this crops up again.

since it seems netspool files are biting you, can you enable networker debug?

you should see lots of NW lines there then.

If you login as aide, do you see many spool tasks?



[#] Fri May 16 2014 10:57:32 EDT from Freakdog @ Dog Pound BBS II

Subject: Re: DB: Panic: Too many open files.

[Reply] [ReplyQuoted] [Headers] [Print]

 

Thu May 15 2014 11:17:19 AM EDT from dothebart @ Uncensored Subject: Re: DB: Panic: Too many open files.

 

Wed May 14 2014 09:26:07 EDT from Freakdog @ Dog Pound BBS II Subject: DB: Panic: Too many open files.

Citadel 8.24...this only cropped up after I had to do a full system restore on the host server and then the VM that runs the Citadel instance. Unfortunately, this seems to be happening daily.

May 14 03:07:17 bbs citserver[26719]: NW[comalies][2991]: polling
May 14 03:07:17 bbs citserver[26719]: NW[uncnsrd][2992]: polling
May 14 03:07:17 bbs citserver[26719]: NW[trada][2993]: polling
May 14 03:07:17 bbs citserver[26719]: serv_extnotify: processing notify queue

I wouldn't imagine that 177MB for cdb.00 or 423MB for cdb.07 should be too large for the DB engine...I'm sure that the DB files on Uncensored must be far larger.

I took the liberty, a couple of days ago, of increasing the nofiles ulimit for the "bbs" user from the default 1024 to 1536, though lsof only shows 417 "REG" entries.

Digging a little deeper, though, I'm seeing that 360 of those lsof hits are for one particular file in my /usr/local/citadel/network/spoolout directory (the spool file for the trada node, and that file is only 721.6K).

I've restarted the citadel and am watching the number of file handles (per lsof) grow, and the only network spool file is trada. Wondering if it's an issue with this particular spoolfile, I'm renaming the file and waiting to see if this crops up again.

since it seems netspool files are biting you, can you enable networker debug?

you should see lots of NW lines there then.

If you login as aide, do you see many spool tasks?

At present, since I moved the netspool file out, the number of file handles hasn't grown above 60, and I have had nary an issue.



[#] Sun May 18 2014 12:57:42 EDT from fastfinge @ Uncensored

Subject: How do I Clear SMTP Outgoing Queue in Citadel Server via command line?

[Reply] [ReplyQuoted] [Headers] [Print]

I had a spam attack this morning. I cleaned up the account responsible, but now I am stuck with over 80 thousand messages in my citadel SMTP queue. This queue is so large that I can't use WebCit to clear the queue, because it just crashes trying to load all the messages to get them all selected and deleted. Surely there must be some command line way to clear this queue that doesn't require 6 different JavaScript libraries? The instructions given in the citadel documentation only mention WebCit, and I can't get that to work at all due to the size of the queue.

Go to page: [1] 2 3 4 5 ... Last