George,

My reply was in reference to DR only.  I have no problems using DBMS for 
instantiated queues.

Richard Tsujimoto 
Consultant 
IT Infrastructure 
Canon U.S.A., Inc. 
( Office:  516-328-5797 
2  Fax:     516-328-4588 
* Email: [EMAIL PROTECTED] 



George Carey <[EMAIL PROTECTED]> 
Sent by: MQSeries List <MQSERIES@LISTSERV.MEDUNIWIEN.AC.AT>
03/30/2008 02:27 PM
Please respond to
MQSeries List <MQSERIES@LISTSERV.MEDUNIWIEN.AC.AT>


To
MQSERIES@LISTSERV.MEDUNIWIEN.AC.AT
cc

Subject
Re: New Topic:  MQSeries some new Persistence Options







What exactly is your point here guys, that having Queue's instantiated in 
Databases is a bad idea ?? I am missing it. 
 
Why are TIBCO, IBM (WAS JQM) and Oracle all giving or have the option of 
instantiating Queues in databases ... for no reason??
 
Sure hardware replication should be used, but why not a blend of hardware 
and software technologies to give the best solution?
 
GTC
-----Original Message-----
From: Robert Broderick [mailto:[EMAIL PROTECTED]
Sent: Sunday, March 30, 2008 08:31 AM
To: MQSERIES@LISTSERV.MEDUNIWIEN.AC.AT
Subject: Re: New Topic: MQSeries some new Persistence Options

yes, agreed. Had a client once that was doing this. With TIBCO, slightly a 
different beast and approach than the one suggested. but all the same the 
architecture idea was the same. Against my wishes and my future job stock 
options prospects I lost the argument and they implemented. A short while 
after I ran out the back door, their system, I heard from a reliable 
source, came to a screeching halt. Hours went by before they could get the 
pig back up. I always like that line.....
 
"Lunacy is doing the same thing twice and expecting different results."




Date: Sat, 29 Mar 2008 11:22:16 -0400
From: [EMAIL PROTECTED]
Subject: Re: New Topic: MQSeries some new Persistence Options
To: MQSERIES@LISTSERV.MEDUNIWIEN.AC.AT


Relying on software-based replication for DR purposes is not a good idea. 
A better solution would be to use storage managed devices that also 
replicate data to remote DR sites.

Richard Tsujimoto 
Consultant 
IT Infrastructure 
Canon U.S.A., Inc. 
( Office:  516-328-5797 
2  Fax:     516-328-4588 
* Email: [EMAIL PROTECTED] 



George Carey <[EMAIL PROTECTED]> 
Sent by: MQSeries List <MQSERIES@LISTSERV.MEDUNIWIEN.AC.AT> 03/28/2008 
04:58 PM 

Please respond to
MQSeries List <MQSERIES@LISTSERV.MEDUNIWIEN.AC.AT>



To
MQSERIES@LISTSERV.MEDUNIWIEN.AC.AT 
cc

Subject
New Topic:  MQSeries some new Persistence Options









Is it time (indeed perhaps overdue) that IBM puts forth new options on 
storage media for persisting messages and logs? 
  
Oracle AQ has from the outset used its database as the storage medium. IBM 
even initially entertained the use of DB2 as its persistence 
medium but opted for its current flat file structure with proprietary 
keying mechanism yet I understand still retained logic taken right from 
DB2 
source for its message logging mechanism. 
  
Currently competitors such as Tibco (EMS) are in Beta with new product 
feature to allow messages to be persisted to Oracle or DB2 
or theoretically any ODBC, JDBC compliant DBMS. It is being targeted for a 
May 2008 release. Just call Tibco they will tell you this!! 
  
One of the advantages of optionally picking a database vendor for 
Queue/message persistence is obviously a robust and completely 
time tested set of solutions (30 years worth) around the many and varied 
issues relating to DR, HA  as well as distributed transactions. 
  
Case in point, you have a need for site to site replication for both DR 
and HA requirements. Your apps send and receive data via ESB 
and messages can at any point in time be in a state of flux holding 10s of 
thousands of messages that need to be inserted or updated 
into sundry database tables. 
  
How do you assure that the message queues and databases are in sync with 
one another on primary and secondary, ..., sites?  It clearly 
is much simpler if the Database and Messaging persistence stores are one 
and the same, a la Oracle and Oracle AQ.  It will soon be the 
case for Oracle and Tibco or DB2 and Tibco, or AnyODBC and Tibco.  What 
about WSMQ and ????. 
  
Maybe WSMQ could keep its legacy persistent store with some added features 
like circular logging with damaged queue recovery capability!!! 
  
But perhaps options for industry standards of ODBC and JDBC databases as 
the persistent stores will be needed to compete.   
  
Also thinking (out of the IBM box) about persistence options, how about 
looking into the use NetApps like devices in conjunction with concurrent 
access features like 
those in the Linux GFS file systems together with some tweaking of the 
WSMQ servers that might even enable the capabilities of shared message 
queues for other than 
the mainframe!! 
  
Just some thoughts!! Any feedback is welcome. 
  
GTC 
 
 

 




List Archive - Manage Your List Settings - Unsubscribe Instructions for 
managing your mailing list subscription are provided in the Listserv 
General Users Guide available at http://www.lsoft.com 



List Archive - Manage Your List Settings - Unsubscribe Instructions for 
managing your mailing list subscription are provided in the Listserv 
General Users Guide available at 
http://www.lsoft.com/resources/manuals.asp 

Test your Star IQ Play now! 


List Archive - Manage Your List Settings - Unsubscribe 
Instructions for managing your mailing list subscription are provided in 
the Listserv General Users Guide available at 
http://www.lsoft.com/resources/manuals.asp 


List Archive - Manage Your List Settings - Unsubscribe 
Instructions for managing your mailing list subscription are provided in 
the Listserv General Users Guide available at http://www.lsoft.com 



To unsubscribe, write to [EMAIL PROTECTED] and,
in the message body (not the subject), write: SIGNOFF MQSERIES
Instructions for managing your mailing list subscription are provided in
the Listserv General Users Guide available at http://www.lsoft.com
Archive: http://listserv.meduniwien.ac.at/archives/mqser-l.html

<<image/gif>>

<<image/gif>>

<<image/gif>>

<<image/gif>>

Attachment: ATT00001
Description: Binary data

Attachment: ATT00002
Description: Binary data

Attachment: ATT00003
Description: Binary data

Attachment: ATT00004
Description: Binary data

Reply via email to