ASG
IBM
Zystems
Cressida
Icon
Netflexity
 
  MQSeries.net
Search  Search       Tech Exchange      Education      Certifications      Library      Info Center      SupportPacs      LinkedIn  Search  Search                                                                   FAQ  FAQ   Usergroups  Usergroups
 
Register  ::  Log in Log in to check your private messages
 
RSS Feed - WebSphere MQ Support RSS Feed - Message Broker Support

MQSeries.net Forum Index » Clustering » cluster sender - can't seem to change conname

Post new topic  Reply to topic
 cluster sender - can't seem to change conname « View previous topic :: View next topic » 
Author Message
dingdong
PostPosted: Thu Feb 19, 2009 12:19 pm    Post subject: cluster sender - can't seem to change conname Reply with quote

Novice

Joined: 12 Nov 2008
Posts: 14

I had to change the IP address on a server that is running two queue managers in a cluster. I'm having trouble with the cluster sender channel starting. It's RedHat Linux - ES 4 Update 4. MQ 6.0.2.5.

The two queue managers are:

QJAOMSD1 (full repository)
JAOTBR01

The cluster receiver channel def on QJAOMSD1:

Code:
dis chl (TO.QJAOMSD1)
     1 : dis chl (TO.QJAOMSD1)
AMQ8414: Display Channel details.
   CHANNEL(TO.QJAOMSD1)                    CHLTYPE(CLUSRCVR)
   ALTDATE(2009-02-19)                     ALTTIME(18.51.18)
   BATCHHB(0)                              BATCHINT(0)
   BATCHSZ(50)                             CLUSNL( )
   CLUSTER(OMHTCLUS)                       CLWLPRTY(0)
   CLWLRANK(0)                             CLWLWGHT(50)
   COMPHDR(NONE)                           COMPMSG(NONE)
   CONNAME(omhwmb1qa.qa2.gmsmail.com(1414))
   CONVERT(NO)                             DESCR( )
   DISCINT(6000)                           HBINT(300)
   KAINT(AUTO)                             LOCLADDR( )
   LONGRTY(999999999)                      LONGTMR(1200)
   MAXMSGL(4194304)                        MCANAME( )
   MCATYPE(THREAD)                         MCAUSER( )
   MODENAME( )                             MONCHL(QMGR)
   MRDATA( )                               MREXIT( )
   MRRTY(10)                               MRTMR(1000)
   MSGDATA( )                              MSGEXIT( )
   NETPRTY(0)                              NPMSPEED(FAST)
   PUTAUT(DEF)                             RCVDATA( )
   RCVEXIT( )                              SCYDATA( )
   SCYEXIT( )                              SENDDATA( )
   SENDEXIT( )                             SEQWRAP(999999999)
   SHORTRTY(10)                            SHORTTMR(60)
   SSLCAUTH(REQUIRED)                      SSLCIPH( )
   SSLPEER( )                              STATCHL(QMGR)
   TPNAME( )                               TRPTYPE(TCP)



The cluster sender channel def on JAOTBR01:

Code:
dis chl (TO.QJAOMSD1)
     1 : dis chl (TO.QJAOMSD1)
AMQ8414: Display Channel details.
   CHANNEL(TO.QJAOMSD1)                    CHLTYPE(CLUSSDR)
   ALTDATE(2009-02-19)                     ALTTIME(18.51.18)
   BATCHHB(0)                              BATCHINT(0)
   BATCHSZ(50)                             CLUSNL( )
   CLUSTER(OMHTCLUS)                       CLWLPRTY(0)
   CLWLRANK(0)                             CLWLWGHT(50)
   COMPHDR(NONE)                           COMPMSG(NONE)
   CONNAME(omhwmb1qa.qa2.gmsmail.com(1414))
   CONVERT(NO)                             DESCR( )
   DISCINT(6000)                           HBINT(300)
   KAINT(AUTO)                             LOCLADDR( )
   LONGRTY(999999999)                      LONGTMR(1200)
   MAXMSGL(4194304)                        MCANAME( )
   MCATYPE(THREAD)                         MCAUSER( )
   MODENAME( )                             MONCHL(QMGR)
   MSGDATA( )                              MSGEXIT( )
   NPMSPEED(FAST)                          PASSWORD( )
   RCVDATA( )                              RCVEXIT( )
   SCYDATA( )                              SCYEXIT( )
   SENDDATA( )                             SENDEXIT( )
   SEQWRAP(999999999)                      SHORTRTY(10)
   SHORTTMR(60)                            SSLCIPH( )
   SSLPEER( )                              STATCHL(QMGR)
   TPNAME( )                               TRPTYPE(TCP)
   USERID( )



Channel status of TO.QJAOMSD1 (cluster sender) shows the OLD IP address in CONNAME:

Code:
dis chs (TO.QJAOMSD1)
     1 : dis chs (TO.QJAOMSD1)
AMQ8417: Display Channel Status details.
   CHANNEL(TO.QJAOMSD1)                    CHLTYPE(CLUSSDR)
   CONNAME(57.4.3.49(1414))                CURRENT
   RQMNAME( )                              STATUS(RETRYING)
   SUBSTATE( )                             XMITQ(SYSTEM.CLUSTER.TRANSMIT.QUEUE)


And I see the following in the error log:

Code:
02/19/2009 07:39:19 PM - Process(32110.18) User(wmbadmin) Program(amqrmppa)
AMQ9202: Remote host 'omhtest (57.4.3.49) (1414)' not available, retry later.

EXPLANATION:
The attempt to allocate a conversation using TCP/IP to host 'omhtest
(57.4.3.49) (1414)' was not successful.  However the error may be a transitory one and it may be possible to successfully allocate a TCP/IP conversation later.


The new IP address should be 57.4.26.195:

Code:
[wmbadmin@omhwmb1qa errors]$ dig omhwmb1qa.qa2.gmsmail.com +short
57.4.26.195


And it's definitely listening:

Code:
[wmbadmin@omhwmb1qa ~]$ telnet omhwmb1qa.qa2.gmsmail.com 1414
Trying 57.4.26.195...
Connected to omhwmb1qa.qa2.gmsmail.com (57.4.26.195).
Escape character is '^]'.
^]
telnet> q
Connection closed.


I've stopped/started the channels and both queue managers, but still the same behavior.

If someone could please tell me what I'm missing, I'd appreciate it.

Tks/Kim
Back to top
View user's profile Send private message
vandi
PostPosted: Thu Feb 19, 2009 12:49 pm    Post subject: Reply with quote

Acolyte

Joined: 13 Dec 2008
Posts: 67

Did you Alter the existing channel to take new Host name or deleted the existing(old host) channel and redefined a New channel that points to new Host?


Thanks
Vandi
Back to top
View user's profile Send private message
dingdong
PostPosted: Thu Feb 19, 2009 1:03 pm    Post subject: Reply with quote

Novice

Joined: 12 Nov 2008
Posts: 14

Hi.

Initially, I altered the channel. When that didn't work, I deleted it and recreated it (both sender and receiver).

Tks/Kim
Back to top
View user's profile Send private message
vandi
PostPosted: Thu Feb 19, 2009 1:09 pm    Post subject: Reply with quote

Acolyte

Joined: 13 Dec 2008
Posts: 67

Can you do the telnet from both ends(ie. Partial Repo and Full Repo as well) and paste the otput.
Back to top
View user's profile Send private message
dingdong
PostPosted: Thu Feb 19, 2009 1:19 pm    Post subject: Reply with quote

Novice

Joined: 12 Nov 2008
Posts: 14

Sure.

Here are the telnets:

[root@omhwmb1qa ~]# telnet omhwmb1qa.qa2.gmsmail.com 1414
Trying 57.4.26.195...
Connected to omhwmb1qa.qa2.gmsmail.com (57.4.26.195).
Escape character is '^]'.
^]
telnet> q
Connection closed.

[root@omhwmb1qa ~]# telnet omhwmb1qa.qa2.gmsmail.com 10100
Trying 57.4.26.195...
Connected to omhwmb1qa.qa2.gmsmail.com (57.4.26.195).
Escape character is '^]'.
^]
telnet> q
Connection closed.

And the listener proceses:

[root@omhwmb1qa ~]# ps -ef|grep runmqlsr|grep -v grep
mqm 32024 32002 0 19:29 ? 00:00:00 /opt/mqm/bin/runmqlsr -r -m QJAOMSD1 -t TCP -p 1414 -b 10
mqm 32102 32080 0 19:29 ? 00:00:00 /opt/mqm/bin/runmqlsr -r -m JAOTBR01 -t TCP -p 10100

Tks/Kim
Back to top
View user's profile Send private message
vandi
PostPosted: Thu Feb 19, 2009 1:40 pm    Post subject: Reply with quote

Acolyte

Joined: 13 Dec 2008
Posts: 67

From your results, I think both Full Repository and Partial repository are on the same server.

Did you change the conname for both Queue Managers(Partial & Full Repos)?
Back to top
View user's profile Send private message
dingdong
PostPosted: Thu Feb 19, 2009 1:47 pm    Post subject: Reply with quote

Novice

Joined: 12 Nov 2008
Posts: 14

Hi.

Yes, they are located on the same server. When you say "Did you change the conname for both Queue Managers(Partial & Full Repos)?", you are talking about the channels, correct? If so, yes. I had even tried using just "localhost", but it still keeps showing the old IP address (57.4.3.49) in the error logs and on the channel status.

Tks/Kim
Back to top
View user's profile Send private message
Sam Uppu
PostPosted: Thu Feb 19, 2009 2:00 pm    Post subject: Reply with quote

Yatiri

Joined: 11 Nov 2008
Posts: 610

How did you change the host?. Did you stop the QMgrs and moved the /var/mqm filesystem to the new server?.

When you are in Full repository (runmqsc FR_QMgr)

Please check this:

DIS CLUSQMGR(PR_QMGR) QMID

If you still see, the old PR_QMgr with old QMID(creation date & timestamp), please remove it with

RESET CLUSTER(CLUSTERNAME) ACTION(FORCEREMOVE) QMID(OLD PR_QMID) QUEUES(YES)

Execute the above command until all the older information got removed.

Then,
Come onto PR QMgr(runmqsc PR_QMGR)

DIS CLUSQMGR(FR_QMGR) QMID

If you still see, the old FR_QMgr with old QMID(creation date & timestamp), please remove it with

RESET CLUSTER(CLUSTERNAME) ACTION(FORCEREMOVE) QMID(OLD FR_QMID) QUEUES(YES)


Then establish the new cluster sender/receiver channels on both ends(FR & PR)

Thanks.
Back to top
View user's profile Send private message
fjb_saper
PostPosted: Thu Feb 19, 2009 3:07 pm    Post subject: Reply with quote

Grand High Poobah

Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY

dingdong wrote:
Hi.

Yes, they are located on the same server. When you say "Did you change the conname for both Queue Managers(Partial & Full Repos)?", you are talking about the channels, correct? If so, yes. I had even tried using just "localhost", but it still keeps showing the old IP address (57.4.3.49) in the error logs and on the channel status.

Tks/Kim

You need to stop and restart the channel for it to pick up the change.
A cluster sender (not to an FR) is changed by changing the corresponding cluster receiver and bouncing the cluster receiver. Have fun
_________________
MQ & Broker admin
Back to top
View user's profile Send private message Send e-mail
exerk
PostPosted: Fri Feb 20, 2009 1:50 am    Post subject: Reply with quote

Jedi Council

Joined: 02 Nov 2006
Posts: 6339

Sam Uppu wrote:
How did you change the host?. Did you stop the QMgrs and moved the /var/mqm filesystem to the new server?.


Where did that question come from?

The poster stated:

Quote:
I had to change the IP address on a server that is running two queue managers in a cluster...


As regards the rest of your advice:

Sam Uppu wrote:
When you are in Full repository (runmqsc FR_QMgr)

Please check this:

DIS CLUSQMGR(PR_QMGR) QMID

If you still see, the old PR_QMgr with old QMID(creation date & timestamp), please remove it with

RESET CLUSTER(CLUSTERNAME) ACTION(FORCEREMOVE) QMID(OLD PR_QMID) QUEUES(YES)

Execute the above command until all the older information got removed.

Then,
Come onto PR QMgr(runmqsc PR_QMGR)

DIS CLUSQMGR(FR_QMGR) QMID

If you still see, the old FR_QMgr with old QMID(creation date & timestamp), please remove it with

RESET CLUSTER(CLUSTERNAME) ACTION(FORCEREMOVE) QMID(OLD FR_QMID) QUEUES(YES)


Then establish the new cluster sender/receiver channels on both ends(FR & PR)

Thanks.


Very bad advice and totally irrelevant in the context of the posters problems, and in clusters it can take time for the effects of a command to propagate around a cluster, so:

Quote:
Execute the above command until all the older information got removed.'


Is not just bad advice, it is also BAD PRACTICE! You're obviously one of 'those' people that cause me more work than is necessary, the only benefit being that 'those' people keep me in work.

PLEASE read the posts more carefully before offering such spurious advice
_________________
It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys.
Back to top
View user's profile Send private message
Sam Uppu
PostPosted: Fri Feb 20, 2009 7:22 am    Post subject: Reply with quote

Yatiri

Joined: 11 Nov 2008
Posts: 610

exerk wrote:
Sam Uppu wrote:
How did you change the host?. Did you stop the QMgrs and moved the /var/mqm filesystem to the new server?.


Where did that question come from?

The poster stated:

Quote:
I had to change the IP address on a server that is running two queue managers in a cluster...


As regards the rest of your advice:

Sam Uppu wrote:
When you are in Full repository (runmqsc FR_QMgr)

Please check this:

DIS CLUSQMGR(PR_QMGR) QMID

If you still see, the old PR_QMgr with old QMID(creation date & timestamp), please remove it with

RESET CLUSTER(CLUSTERNAME) ACTION(FORCEREMOVE) QMID(OLD PR_QMID) QUEUES(YES)

Execute the above command until all the older information got removed.

Then,
Come onto PR QMgr(runmqsc PR_QMGR)

DIS CLUSQMGR(FR_QMGR) QMID

If you still see, the old FR_QMgr with old QMID(creation date & timestamp), please remove it with

RESET CLUSTER(CLUSTERNAME) ACTION(FORCEREMOVE) QMID(OLD FR_QMID) QUEUES(YES)


Then establish the new cluster sender/receiver channels on both ends(FR & PR)

Thanks.


Very bad advice and totally irrelevant in the context of the posters problems, and in clusters it can take time for the effects of a command to propagate around a cluster, so:

Quote:
Execute the above command until all the older information got removed.'


Is not just bad advice, it is also BAD PRACTICE! You're obviously one of 'those' people that cause me more work than is necessary, the only benefit being that 'those' people keep me in work.

PLEASE read the posts more carefully before offering such spurious advice


Well. I agree. As per Dingdong, the channels are defined with correct CONNAME and the listeners are running. As far as the channels are properly defined with proper conname and listeners are up and running, the channels should be running. Dingdong tried everything like delete and recreate the channels and listeners are running. With this it should work. As both QMgrs are running on the same server there wont be any network connectivity(telnet is working fine).


Coming onto RESET CLUSTER(*)...

I also had the similar situation. I had 2 FRs and 2 PRs. The cluster was working fine(I was able to send the messages across).

Few days back I had this issue: the messages were not going from one of the FR to a PR though nothing has changed. When I do a DIS CHS(TO.PR) on FR, it was showing in retrying state where as the channels are running from PR QMgr side.

On the FR when I do a DIS CLUSQMGR(PR) QMID QMTYPE SUSPEND STATUS, I see many PRs with different QMIDs. Some are older QMIDs which are due to recreation of PR as we had some issues which built from saveqmgr backup.

I tried many ways like Dingdong did...like delete the channels are recreating, rebuild from saveqmgr backup etc. still no success. Then I deleted the older PR information from FR with the above command[RESET CLUSTER(*) QMID ACTION(FORCEREMOVE) QUEUES(YES)].

Then it worked for me.

As per the QMgrs in Cluster Redbook there is nothing harm in removing the older information if anything exists as such. Please check the QMgrs in Cluster Redbook, troubleshooting part page 167.

As everybody is aware that with the channels are properly defined with proper conname and listeners are running on both ends both the channels on both ends should be running when they are defined on one physical server. Then why the channel is is in retry state and why it still showing older information(older IP)?. Do you the answer?. Do you have any other suggestion.

I shared what I know and what I experienced. It may work in your case or may not. Its upto you whether you give it a try or not.

Thanks.
Back to top
View user's profile Send private message
exerk
PostPosted: Fri Feb 20, 2009 7:31 am    Post subject: Reply with quote

Jedi Council

Joined: 02 Nov 2006
Posts: 6339

Sam Uppu wrote:
...Some are older QMIDs which are due to recreation of PR as we had some issues which built from saveqmgr backup...


Some? How many times have you had to rebuild the queue manager? And why wasn't it cleaned up at the time? More importantly, why have you had to rebuild the PR? The rebuild may have removed the symptom, but what was the cause?

Sam Uppu wrote:
...Then why the channel is is in retry state and why it still showing older information(older IP)?...


Because as fjb_saper pointed out: 'You need to stop and restart the channel for it to pick up the change'
_________________
It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys.
Back to top
View user's profile Send private message
Sam Uppu
PostPosted: Fri Feb 20, 2009 7:43 am    Post subject: Reply with quote

Yatiri

Joined: 11 Nov 2008
Posts: 610

exerk wrote:
Sam Uppu wrote:
...Some are older QMIDs which are due to recreation of PR as we had some issues which built from saveqmgr backup...


Quote:
Some? How many times have you had to rebuild the queue manager? And why wasn't it cleaned up at the time? More importantly, why have you had to rebuild the PR? The rebuild may have removed the symptom, but what was the cause?


We had 3 QMIDs(one is latest and 2 from older). Yes, I agree. We no need to remove/ rebuild PR. As we rebuilt it twice in the past, that older PR information will retain for around 90 days in FR and will be removed later on. Thatswhat the Cluster QMgrs redbook is saying. Thatswhy we didn't clean them up as there wont be any harm with that older PR information in FR.

Sam Uppu wrote:
...Then why the channel is is in retry state and why it still showing older information(older IP)?...


Because as fjb_saper pointed out: 'You need to stop and restart the channel for it to pick up the change'


I agree. Dindong tried everything like stop/start, even deleted the channels and recreated... still the same problem. Not sure why the channel status still show older information(older ip).
Back to top
View user's profile Send private message
bruce2359
PostPosted: Fri Feb 20, 2009 8:10 am    Post subject: Reply with quote

Poobah

Joined: 05 Jan 2008
Posts: 9469
Location: US: west coast, almost. Otherwise, enroute.

I seem to recall the WMQ Clusters manual recommending
. first making the object unavailable (stop channel, put-disable a queue),
. then taking the object out of the cluster,
. then altering the object attributes,
. then putting it back into the cluster.
_________________
I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live.
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic  Reply to topic Page 1 of 1

MQSeries.net Forum Index » Clustering » cluster sender - can't seem to change conname
Jump to:  



You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
Protected by Anti-Spam ACP
 
 


Theme by Dustin Baccetti
Powered by phpBB © 2001, 2002 phpBB Group

Copyright © MQSeries.net. All rights reserved.