Local Delivery - Partially Aborted - 5089

[B.O.F.H.]

Hi Community,

I got problems with my Local Delivery Service. since 4 Days, mails will be recieved by fetchmail and delivered to the Local Delivery Service, but the Servcie is Aborted.

I tried to omoff, omreset, the Service. Everytime the Service is instant Aborted.
Ther are 5089 Msgs in the Queue, 5000 are crap notifications from different services to a special mailaccount in scalix. But 89 Msgs are very important! I don't know, how I can reactivate the Service, omresub don't works.

Has anyone an idea?

Sincerly,
Thorsten

[B.O.F.H.]

It's me again, think i got the problem by it's roots.

Notification account has 65532 Msgs in Waste-Basket and Inbox.
Login via Webmail isn't possible, so I used omtidyu -command to cleanup the waste basket and the inbox.

After cleanup of the waste basket and reboot the Local Delivery Service will run, but the Mails in the Delivery Queue will not be delivered.

Is there a way to give the Delivery Service a "little kick" to deliver the msgs?

Greetingz,
Thorsten

[B.O.F.H.]

Hi, me again.

Got the same problem with another Server, on an other location. after 413 Msgs in Local Delivery Queue I stopped the POP3 Daemon fetchmail, with collects the mails form the ISP.

After one nIght the Queue was empty, but the omqdump shows me the following:

Code: Select all

QUEUE         Messages     Active       Ever       Load [1min,5min,15min]
ARCHERR              0          0          0       0.00       0.00       0.00
...
ERROR                1          0          0       1.00       1.00       1.00
LICENSE              0          0          0       0.00       0.00       0.00
LOCAL                0          0        413       0.00       0.00       0.00
...
ROUTER               0          0       1193       0.00       0.00       0.00
..
UNIX                 0          0       1154       0.00       0.00       0.00

Special queues.
IDEL                 0          0       1567       0.00       0.00       0.00
POISON               0          0          0       0.00       0.00       0.00


What means the column "Ever" ??? Is there a chance, that I can extract the mails on any way? or are they lost forever?

Help!!

[Valerion]

The Messages coloumn is the number of messages currently on the queue (being processed and awaiting a queue handler). The Active coloumn indicate how many messages are currently being processed. This should be 1, unless you configured multiple instances of a specific service (which is possible, but adds load to the machine). Ever is the largest value that Messages has been before. In other words, somewhere in the past (since the last time you restarted the server) there was 1193 messages in the ROUTER queue.

The last 3 coloumns, marked Load, is similar to the output you get if you do top under Linux, and gives you an indication of how busy your queue hander is. You can use this to determine if another instance is needed to share load, for example.

So there's no way to retrieve those messages, what you see in the last 4 coloumns are merely historic information to help you plan your server and manage the load on it.

[B.O.F.H.]

Thanks alot for your reply, but means this, the Mails are delivered? or are they lost in digital nirvana?
Is there a log file, that indicates wheter a mail is delviered or not? the audit is not very helpful in this case...

And why comes this message? :

Code: Select all

server:/opt/scalix/diag # omqdump -s

WARNING: This is a diagnostic tool for use by Scalix trained personnel.
   Improper use can cause serious damage.
   If you do not wish to continue, hit return now.



After this message, the process hungs up it seems... after pressing enter for serveral times there is an error msg:

Code: Select all

omqdump : [OM 11795] Access Denied!


On my other server i don't got something like this... instead the Statistik like the post above will be displayed instant.

Thanks,
Thorsten

[Valerion]

You're getting that error because you need a password to enter omqdump and other tools (including omcontain). It is possible to get Scalix not to ask for a password, though, which is probably the case on your other server. The idea is not to keep people out, but to serve as a reminder that these are potentially dangerous (as in mailstore-corrupting) tools, though extremely handy to fix errors.

The messages in the "Ever" coloumn have been processed, otherwise they would go to another queue (POISON, ERROR or similar), or stay in the queue. So you can assume they have been delivered.

[B.O.F.H.]

Thanks again for your reply!

A stone like the himalaya drops from my shoulders... :D

But...
The password for omcontain is removed! There couldn't be a password-question any longer. (removed the line in the configfile)

I removed it on both servers. the first server, let me in without any questions. but the second (with the 5000Msgs queue) don't let me in and shows me the warning message.

I'm a little bit confused.

Greetingz,
Thorsten

[Valerion]

in my /var/opt/scalix/sys/pam.d/omqdump (Scalix 10 server) I removed

Code: Select all

auth required om_diag

and it works fine for me. The file now contains

Code: Select all

auth optional om_unix2om
auth required om_admin
account required pam_permit

Mmmm .. wonder if that should be an auth required om_unix2om instead ... will have to experiment with that on my demo server.

Also note that each utility has its own file in pam.d, so you can remove it for one and not the other.

[B.O.F.H.]

We still working with scalix 9.1 :roll:

I removed the line

auth required om_diag

from /var/opt/scalix/sys/pam.d/omcontain

and on one system it works fine, on the other i don't works...

Very curious...

[mikethebike]

Are you getting any errors in the fatal log (/var/opt/openmail/logs/fatal)?
If you are getting "conatiner extension full" or suchlike, it could be that clearing out the offendng mailbox was not enough.

Run an "omscan -A -U <username> -vfx" on that account.

The stop local delivery, make sure all the PIDs have dies (or kill them) and restart local delivery.

Mick