| Author |
Message
|
| yashgt |
 Posted: Tue Sep 29, 2015 2:19 am Post subject: Preserving MessageID in Pub-sub |
 |
|
Apprentice
Joined: 11 Nov 2014
Posts: 38
|
Hi,
A message is PUT on an Alias Queue with a Message ID say "M1". There is a Topic on which this message is posted by the Alias Queue. There is a subscription which takes the message from the Topic and puts on a local Queue. The Message ID on the local queue is no M1. It is a string of bytes. Also the correlation ID of the message is not M1.
Is it possible to retain the original Message ID of the message that was dropped on the Alias Queue in any of the header fields of the message on the target queue?
Thanks,
Yash |
|
| Back to top |
|
 |
| Vitor |
| Posted: Tue Sep 29, 2015 3:33 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
A message id can never be 'M1' - it's a field of 24 bytes and no kind of string. Any design that tries to use it like a string will discover this the hard way - as you have.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| yashgt |
| Posted: Tue Sep 29, 2015 4:08 am Post subject: |
|
|
Apprentice
Joined: 11 Nov 2014
Posts: 38
|
M1 was just an example. Actually the message is inserted on a Remote Q which is mapped to an Alias Queue. The message ID is autogenerated as a byte array.
RQ -> Tx Q -> Alias Q -> Topic -> Local Q1.
Main question is, how do I preserve the Message ID that go created on RQ all the way to the Local Q1?
In our design we need the Topic because we want the message to reach multiple consumers. Local Q1 is just one consumer. |
|
| Back to top |
|
|
| zpat |
| Posted: Tue Sep 29, 2015 7:34 am Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5849
Location: UK
|
Using topic alias for publication will generate new msgid (and correlid) values unfortunately.
I think I raised an RFE to allow preservation - however I think IBM don't really look at RFEs because I have had some open for years with no progress.
_________________
Well, I don't think there is any question about it. It can only be attributable to human error. This sort of thing has cropped up before, and it has always been due to human error. |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Tue Sep 29, 2015 11:40 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
The problem with preserving the message ID over Pub/Sub is that you should only ever have a single instance of the MSG ID in your system. If you do a pub/sub operation the sender/publisher can't know how many recipients there are 'subscribed' to the Topic.
I'd like to see expiry times preserved over a pub/sub but that is another matter altogether.
_________________
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 |
|
|
| zpat |
| Posted: Tue Sep 29, 2015 10:10 pm Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5849
Location: UK
|
There are such things as options, that IBM could allow to be set on the subscription.
It's in my RFE which is gathering dust like all the others.
_________________
Well, I don't think there is any question about it. It can only be attributable to human error. This sort of thing has cropped up before, and it has always been due to human error. |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed Sep 30, 2015 2:18 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| zpat wrote: |
| It's in my RFE which is gathering dust like all the others. |
How many votes does it have?
An RFE with a small number of votes will languish behind those with larger numbers and more apparent support in the community. This came up in the experts session at MQTC yesterday so I'm speaking with the authority of having heard IBM people say that.
Post the link here and get some support. As someone said yesterday in the same session, "it's the squeaky wheel that gets greased"
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| zpat |
| Posted: Wed Sep 30, 2015 3:41 am Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5849
Location: UK
|
I know what they say, but very few people are interested in voting on any RFE, however that doesn't mean they would not find the option useful/essential if they later discover this feature.
IBM seem to like getting 90% of the way to a really useful feature and then stopping just short of it being of practical value, for the sake of just a small amount of coding.
Given the battle between MQ and freeware (e.g. cloud software) that is beginning to be fought inside corporations, IBM cannot afford this complacency - they need to demonstrate why paying for middleware is better, by delivering on more customers RFEs before MQ becomes seen as a legacy product.
_________________
Well, I don't think there is any question about it. It can only be attributable to human error. This sort of thing has cropped up before, and it has always been due to human error. |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed Sep 30, 2015 7:13 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| zpat wrote: |
| I know what they say, but very few people are interested in voting on any RFE, however that doesn't mean they would not find the option useful/essential if they later discover this feature. |
If very few people are interested in voting, it should be easier to get the most popular RFE....
But seriously, if not this mechanism then what? No software house will spend time and money developing a feature (even a user suggested one) without some concrete evidence that more people will consider it useful than one customer and all the developers that want to implement it.
Not saying that every single customer is not important, nor that your RFE is without merit. Attempting to make a general point here.
It was worse before RFEs
IBM seem to like getting 90% of the way to a really useful feature and then stopping just short of it being of practical value, for the sake of just a small amount of coding.
| zpatt wrote: |
| delivering on more customers RFEs before MQ becomes seen as a legacy product. |
Respectful I think they can make the value proposition from internal development. What we're discussing here is prioritizing that development.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| zpat |
| Posted: Wed Sep 30, 2015 7:41 am Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5849
Location: UK
|
Very few of the RFEs are being delivered from what I can tell.
Unless hundreds clamour for it - the development priority seems to be more on what features marketing want to see (especially in IIB), rather than tidying up existing features to make them really useful.
I can think of many new features that I have found no practical use for since they are so awkward to use. MQ statistics and accounting for example.
IBM documentation expects people to code programs to analyse the event messages - who on earth has time/skills to do that? (Yes, I know there are now some support pacs - but the point remains that IBM labs assume a lot of customer skill/time that doesn't exist), or they expect customers to buy third-party products to finish the job off properly.
Is there a (GUI) message editor in MQ - after 19 years? No.
_________________
Well, I don't think there is any question about it. It can only be attributable to human error. This sort of thing has cropped up before, and it has always been due to human error. |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed Sep 30, 2015 8:46 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
It was also suggested at MQTC that IBM should explicitly list what RFEs are in a given fix pack or release. I think this will help add credibility to the RFE process, or prove your contention it's nothing but a sop with no relevance.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Wed Sep 30, 2015 9:25 am Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7717
|
I've had a few of my RFEs delivered. Now its debatable if the feature was gonna happen anyway, but at least I got some thru.
If you want to be disenchanted with the RFE process, open up a couple for DataPower. Most of those I opened got rejected with some sort of comment "That's not how it works today - rejected." Well duh, what part of "request" and "enhancement " don't they get?
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Wed Sep 30, 2015 3:24 pm Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
| zpat wrote: |
I
Given the battle between MQ and freeware (e.g. cloud software) that is beginning to be fought inside corporations, IBM cannot afford this complacency - they need to demonstrate why paying for middleware is better, by delivering on more customers RFEs before MQ becomes seen as a legacy product. |
I had this discussion with a produt manager the other day. He'd just had a big outage because of a poison message in MSMQ. {don't need to say much more}
'But MSMQ is free' etc etc
'Products like Active MQ are free'....
Then I demoed how we dealt with bad messages.
His exact words were
'And that is in the core product? We don't have to pay extra for it?'
'Yep it is in the core product'
Then I got onto message encryption and how it is done with MQ. he was very impressed.
Now we will have to wait to see if he can convince his developers (C-Sharp devs) that moving to MQ is a good idea. Judging by their reactions I think we are in for an uphill task.
_________________
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 |
|
|
| mqjeff |
| Posted: Thu Oct 01, 2015 4:59 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
Just show them the MQ .NET API and let them play.
Shouldn't be too much issue changing their minds.
_________________
chmod -R ugo-wx / |
|
| Back to top |
|
|
| vicks_mq |
| Posted: Thu Dec 06, 2018 12:53 pm Post subject: |
|
|
Disciple
Joined: 03 Oct 2017
Posts: 162
|
Just wondering if IBM has implemented Message ID preservation in higher version.
we have an application putting request message and we want to send the reply to 2 reply queues using the TOPIC subscription.
Our application copy the message and put it in correlation ID and reply back,
so while Subscriber will put message to both destination, will the correlation ID will remain the same? |
|
| Back to top |
|
|
|
|
