| Author |
Message
|
| serpota |
 Posted: Tue Aug 19, 2014 10:54 pm Post subject: since what version HOSTNAME can not be changed |
 |
|
Voyager
Joined: 26 May 2006
Posts: 85
|
Unfortunately, we have noticed that changing HOSTNAME of the host machine destroys MQ.
I would say this was not happening at MQ version 5 and also version 6.
Is this feature related to MQ version 7 ?
And the main question is ...
What is the reason to include HOSTNAME in "MQ id" ?
What benefit do we get ?
Ramon. |
|
| Back to top |
|
 |
| smdavies99 |
| Posted: Wed Aug 20, 2014 12:17 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
What do you mean 'destroys MQ'?
What platform is this on. also what is the full version and fixpack of MQ you are running?
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Wed Aug 20, 2014 4:01 am Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
And what is "MQ id"?
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Wed Aug 20, 2014 5:26 am Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
I have the sad feeling he is talking about Windows and local ids which can be portrayed as hostname\id.
Well the first task would be to make sure the MQ service ids still have the same characteristics, as to permissions and local policies, make sure you can start the MQ Service, and rerun the configure MQ Wizard.
If that does not work, uninstall and reinstall.
Have fun 
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed Aug 20, 2014 5:53 am Post subject: Re: since what version HOSTNAME can not be changed |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
Assuming my worthy associate is right & you're talking about a Windows id in the format "domain\id" where the machine name forms the domain name:
| serpota wrote: |
| What is the reason to include HOSTNAME in "MQ id" ? |
No reason I can think of and, speaking entirely personally, from MQv5 to the present day I've made a point of always running MQ as a local id unqualified in any way. The dangers of using a true domain id to run MQ have been extensively discussed over the years of course
| serpota wrote: |
| What benefit do we get ? |
None that I can think of
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| serpota |
| Posted: Wed Aug 20, 2014 9:01 am Post subject: |
|
|
Voyager
Joined: 26 May 2006
Posts: 85
|
Sorry of my wrong expression.
MQ as product is not destroyed.
But queue managers do not run anymore.
And I dont think it is only in Winodws ... not sure but. |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed Aug 20, 2014 9:12 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| serpota wrote: |
| But queue managers do not run anymore. |
With what messages in the various logs?
| serpota wrote: |
| And I dont think it is only in Winodws |
Then you need to explain more clearly what you mean by:
| serpota wrote: |
| What is the reason to include HOSTNAME in "MQ id" ? |
Because (like my associate) the only platform where the id is anything other than a string literal is Windows, where the id can include a domain qualifier and for a locally defined id the domain is either omitted or is the machine name (hostname).
So if you don't mean that, what do you mean?
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| serpota |
| Posted: Thu Aug 21, 2014 12:34 am Post subject: |
|
|
Voyager
Joined: 26 May 2006
Posts: 85
|
Good morning.
In the windows machine I am using right now I have a quite a complex environment set up which I dont want to destroy : 2 Front End qmgr's, 3 MB server's qmgr's, 2 FR's, plus 2 qmgrs more for aux functions.
Monday I shall have a smaller windows machine available, so I shall change hostname there and tell you the messages.
But basically I remember it said the queue manager could not start (due to ... I dont remember, but something related to mq-id).
If this problem does not happen under Linux, it shall be a definitive advantadge to use Linux in the future and skip windows forever, in my point of view.
So, you are saying that mq-id does not exist under Linux ?
Seems strange to me ...
This is a (windows) typical id :
| Code: |
QMID(QM_CNT_2014-03-01_00.21.36)
|
Thanks. |
|
| Back to top |
|
|
| exerk |
| Posted: Thu Aug 21, 2014 1:21 am Post subject: |
|
|
Jedi Council
Joined: 02 Nov 2006
Posts: 6339
|
The QMID is a product of the queue manager name and date/time of creation, and is the manner of ensuring uniqueness.
If your site standard is to use server names as part of a queue manager name - bad practice as far as I am concerned - then changing the server name will have no effect on the QMID and should not stop the queue manager from starting.
If you are using DNS names within your channel definitions, and you change the server name, you will have to change the channel definitions to reflect that name change, so in my opinion it's always better to have a DNS alias for servers hosting queue managers rather than using the literal server name.
Just my 2 cents...
_________________
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 |
|
|
| Vitor |
| Posted: Thu Aug 21, 2014 4:47 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| serpota wrote: |
| But basically I remember it said the queue manager could not start (due to ... I dont remember, but something related to mq-id). |
Clearly this information will be a massive help to our attempts to advise you.... 
| serpota wrote: |
| If this problem does not happen under Linux, it shall be a definitive advantadge to use Linux in the future and skip windows forever, in my point of view. |
Well in my personal view avoiding Windoze and moving to Linux is a definite advantage under all circumstances but there's a certain level of personal bias in that personal view.
| serpota wrote: |
| So, you are saying that mq-id does not exist under Linux ? |
If by "mq-id" you really mean QMID as shown:
| serpota wrote: |
This is a (windows) typical id :
| Code: |
QMID(QM_CNT_2014-03-01_00.21.36)
|
|
Then it exists and is used by all the distributed versions of WMQ including Linux. (Also on z/OS for the record).
With respect to my worthy associate and his comments the QMID is generated when the queue manager is created and then has no correlation to the outside world. So if you use the server name as all or part of the queue manager name (as we do) and change that name then clearly your queue manager name (and it's QMID) won't now match but that's the only problem as he correctly points out. The queue manager will (or should) start as he says.
Now if the change to server name means a change to DNS, he's right to point out that 1-n channels may not start until the address is corrected. But that's very different from the queue manager not starting.
I look forward to your posting of the error message with interest.
Also a topology with 9 queue managers is not that complex. I have 14 queue managers in my test environment, 28 in each of QA & Prod and I like to think that's nice & straightforward. 
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| serpota |
| Posted: Thu Aug 21, 2014 7:28 am Post subject: |
|
|
Voyager
Joined: 26 May 2006
Posts: 85
|
mr EXERC - in my servers, the IP's are fixes, so changing the HOSTNAME has no effect on DNS
mr VITOR - maybe 9 qmgrs are not complex to you, but creating them, configuring them, creating all the objects, adapting some apps to send messages and answer responses, configure some monitoring app, adapt some cleaning tools, etc etc etc is of course not a matter of minutes, at least for me (yet).
Will destroy a simple MQ next Monday.
Thanks. |
|
| Back to top |
|
|
| exerk |
| Posted: Thu Aug 21, 2014 7:33 am Post subject: |
|
|
Jedi Council
Joined: 02 Nov 2006
Posts: 6339
|
| serpota wrote: |
| Will destroy a simple MQ next Monday. |
Would you please clarify by what you mean by this statement?
_________________
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 |
|
|
| bruce2359 |
| Posted: Thu Aug 21, 2014 7:45 am Post subject: |
|
|
Poobah
Joined: 05 Jan 2008
Posts: 9469
Location: US: west coast, almost. Otherwise, enroute.
|
| serpota wrote: |
| maybe 9 qmgrs are not complex to you, but creating them, configuring them, creating all the objects, adapting some apps to send messages and answer responses, configure some monitoring app, adapt some cleaning tools, etc etc etc is of course not a matter of minutes, at least for me (yet). |
Creating 9 qmgrs should not be nine times harder or nine times more complicated than creating a pair of requester-responder qmgrs. Yes, it's complicated, but it's also mostly redundant activities.
Have you developed o/s command scripts to create qmgrs?
Have you developed scripts of MQSC commands to create you object definitions?
Do you have naming standards for your objects?
_________________
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 |
|
|
| Vitor |
| Posted: Thu Aug 21, 2014 8:21 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| bruce2359 wrote: |
| serpota wrote: |
| maybe 9 qmgrs are not complex to you, but creating them, configuring them, creating all the objects, adapting some apps to send messages and answer responses, configure some monitoring app, adapt some cleaning tools, etc etc etc is of course not a matter of minutes, at least for me (yet). |
Creating 9 qmgrs should not be nine times harder or nine times more complicated than creating a pair of requester-responder qmgrs. Yes, it's complicated, but it's also mostly redundant activities. |
| bruce2359 wrote: |
Have you developed o/s command scripts to create qmgrs?
Have you developed scripts of MQSC commands to create you object definitions? |
If not, then you've no real grounds to complain how long it takes you. If you have and it's still not a matter of minutes, you need better scripts.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| mqjeff |
| Posted: Thu Aug 21, 2014 8:34 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
Look, maybe it's just better to wait until Monday when serpota can give some real data about what's really going wrong.
There shouldn't be any part of any version of MQ that is tightly coupled to the host name of a given machine... unless somehow, maybe, you're running a limited-time demo ediiton.
but that's almost certainly not what serpota is doing. |
|
| Back to top |
|
|
|
|
