Mail Routing issues R6.0 XSP model

HCL Domino Domino Forum
1

We have an intermittent issue with routing mail to external internet domains. Our setup is as follows: All servers are in the same Domino domain, 2 Clustered Domino Mail servers route mail to the internet via 2 Clustered Domino SMTP servers. There is no DNS setup on the internal Mail servers only on the SMTP servers. A Foreign SMTP and SMTP connection documents method is used to route external messages from the Mail servers to the SMTP servers.

Below is a print out from both the Miscellaneous and Mail Events views from the log database of the Mail Server when a message was not routing. As you can see from both views, the router seems to be waiting for a DNS. Any messages for internal users get routed (see 8:40 on the Mail Routing view). After quitting the router and restarting it, the external message then gets routed to the SMTP servers, (look at the end of both views at 8:50)

We have other strange things happening, ie. Connection from server … not used, Server not found (see 8:33 in the Miscellaneous view), public key not found for user(see 8:46 in the Miscellaneous view), also, not in the views below, but sometimes we will get an error message saying user not found, you send the same message again and it works!!. All these errors seem to be as if the names.nsf is loosing part of its indexes intermittently or cache is not refreshed correctly

If you try to rebuild the directory index (load updall names.nsf -R ), the problem appears to compound itself and the only solution is to restart the server.

We have been able to reproduce this problem on the client site as well as test servers in our office (both setup for ASP model).

Your help and thoughts would be gratefully appreciated.

02/09/2003 08:33:50 Router: notes.ini setting for Log_Mailrouting being used (note - this option may now be configured in a Server Configuration document)

02/09/2003 08:33:50 Router: Connection from server ISRVR01/INTSRVR/XXXXXXXXXnot used; Server not found in Domino Directory.

02/09/2003 08:33:50 Router: Connection from server ISRVR01/INTSRVR/

XXXXXXXXX not used; Server not found in Domino Directory.

02/21/2003 09:02:56 Router: Unable to dispatch message 001BBBD5 to domain hotmail.com. DNS is unavailable or query timed out, message will be requeued.

02/21/2003 09:02:56 Router: Unable to dispatch message 001BBBDE to domain hotmail.com. DNS is unavailable or query timed out, message will be requeued.

2

Subject: Mail Routing issues R6.0 XSP model

I had this kind of problems also with this error message and no connection found to other servers and directory assistance not working. I updated to version 6.02 beta and this solved most of my problems. Better wait for the final release.

3

Subject: RE: Mail Routing issues R6.0 XSP model

Have installed 602cf1 this din’t solved the router problems. Workarround create a program doc that stops the router task and a program doc that starts the router task 1 min later. This reduces the problem but still having mail not delivered

4

Subject: Latest update and findings regarding this issue

We have noticed that there 2 types of problems with the mail routing whichappear from time to time.

  1. The mail router at time stops sending email while it’s waiting for DNS

availability. The servers have been configured to send all SMTP emails to

another server for routing and therefore should not require any DNS

resolution. This problem is intermittent and hard to reproduce.

  1. The second problem is due to index update of the Domino Directory (see

below for further details). After adding/updating number of users/groups

or other configuration documents in the Directory, the indexes on the

server are update (see the log files). Although this happens very often,

at times it causes the router to loose track of the connection and server

documents (See the errors from ISRVR01 on 22/02/2003 15:46). If the server

is restarted, the error disappears. As you will see in the log file this

error does appear after every update.

The next problem is more generic however it’s the reason behind the above

problem (number 2).

  1. The above problem with indexes not only causes mail routing problems,

but also creates other problems such as user unable to login with the error

message “Your public key is not found in the Domino Directory”.

As you can see the constant update of the directory indexes seem to cause

the problem. If we do not make any modification to the Directory for

number of days these error (2 & 3 ) do not appear.

One way to re-produce the above problems (2 & 3) is to run “updall -R” on

the names.nsf while the server is running and you will all kinds of error

appear. This is true in the case of R6.0 on AIX and Win32 & R6.0.1 on

Win32.

Anyone else facing this problem?

Thomas Larsen73

5

Subject: You are not alone

Regarding the routing problem “user not listed in Domino Directory” in xSP mode, there are several postings here, but no solutions. (For example by Tommy Tähkänen, David I Lazaroff, Richard E Cooke).I ran into the same problem and so my base org (i.e. the Hosting Org) can not receive mail after mail for a hosted org has been received. Since we are only a few people in the hosting org I have simply set up a hosted org that mirrors (manually … ) the hosting org and delivers the mail. But this is hardly satisfactory, maybe someone from Lotus can comment if this is a bug that will be fixed.

regards

6

Subject: RE: You are not alone

HI KaiI got some feed back from the labs and they recommend to enable some bugging variables

The views in the Names & Address Books)NABs) apprear to be out of sync.

Also the indexes in the NABs appear wrong/incorrect.

For SET to better identify potential corruption, please try the following

test in a test environment similar to ISRV01, ISRV02;

1: Shutdown the Domino Server.

2: Run the following commands on the NAB at the OS level;

FIXUP, updall -r, compact -d.

3: Then enable in the Server notes.ini the following debug notes.ini

paramaters to help identify/capture the root cause of the problem:

Log_View_Events=1

LOG_UPDATE=2

DEBUG_TRAP_CORRUPTION=1

4: Restart the Domino Server.

5: Perform the steps to reproduce the problems occuring specific to

incident 1558312 (Domino router not working as expected).

(eg.simulating multiple user registrations, etc )

6: Instead of the normal workaround of restarting the router, try the

following server console command first;

tell router update config

This is to confirm if the routing tables are corrupt, as opposed to

corruption problems with the NAB.

If the above command fails to resolve the problem, run the usual tell

router restart command.

See you can apply the same and provide the passport advantage if you have escalated a call.

thomas

7

Subject: RE: You are not alone

Thanks for the advice thomas. since my server is at the moment in production and the workaround is working I cannot debug just now. Soon as I find time I will research this further in a test environment and let you know.

regards

8

Subject: Well, you are really not alone!

Hello Thomas, hello Kai,

i’ve got the same problem as you and may other in this forum. I’ve tried the suggestions Thomas mentioned…

fixup, updall, compact

and the debug parameters. Additionally i used the debug parameter DEBUGROUTERLOOKUP=3. With this debug option i’ve found out that the name who should be looked up has been replaced with a variable %a (see in the log below). At this time i haven’t a explanation for this!

I would be very, very glad if you maybe have already a solution for this problem or if my explanation could perhaps help you knowing whats wrong.

(Sorry for my bad english g)

Thank You

Markus

27.03.2003 16:04:27 SMTP Server: natsmtp00.webmailer.de (192.67.198.74) connected

27.03.2003 16:04:27 SMTP Server: Message 0052CE4E (MessageID: 008801c2f472$2be56c20$3301a8c0@fundeldv) received

27.03.2003 16:04:27 SMTP Server: natsmtp00.webmailer.de (192.67.198.74) disconnected. 1 message[s] received

27.03.2003 16:04:28 DebugRLookup: Lookup beginning for user: markusfundel name: markusfundel@fundeldv.de

27.03.2003 16:04:28 DebugRLookup: Lookup error: UserName: markusfundel Error: User %a not listed in Domino Directory

27.03.2003 16:04:28 Router: Unable to deliver message 0052CE4E to markusfundel@fundeldv.de

27.03.2003 16:04:28 User markusfundel@fundeldv.de not listed in Domino Directory

9

Subject: same thing here here, almost a workaround !

we are running a R6 Farm in xSP mode, and have the same problemthe only way on limitting undelivery is to restart automaticaly the router task regularly

we have set a program doc, that restart the router every hour, we have much less delivery failure

hope it helps, while waiting the fixes

10

Subject: RE: same thing here here, almost a workaround !

just found that registering users whit Notes ID works fine, the others whitout have problems to get their mail, a workaround seems to be re-register the users and mark Create Notes ID

11

Subject: Mail Routing issues R6.0 XSP model

I’m having the same problem with the SMTP server giving the error: “DNS is unavailable or query timed out, message will be requeued.”

It just started out of the blue. I’ve tried all kinds of things and nothing seems to remedy the problem. All the internet mail is stuck on the mail.box and won’t route out to the internet or be passed on to the notes user.

In the config doc I disabled reverse DNS lookup for senders…and some mail started coming through…but still getting a lot of the other errors.

Does anyone have any other ideas or solutions which have worked for them?

Thanx!

Sincerely, Gurumustuk Singh Khalsa

=========================================

SikhNet - http://www.sikhnet.com

=========================================