Not that I've ever found. There is a procedure for advancing the Redbook version numbers but it involves putting a bunch of MQ experts in a test lab and feeding them pizza and caffeine for a month.
-- T.Rob -----Original Message----- From: Conklin, William [mailto:[EMAIL PROTECTED]] Sent: Thursday, December 12, 2002 2:33 PM To: [EMAIL PROTECTED] Subject: Re: What events cause a media image to be taken? This manual is dated 1998 is there a newer version? -----Original Message----- From: Wyatt, T. Rob [mailto:[EMAIL PROTECTED]] Sent: Thursday, December 12, 2002 1:02 PM To: [EMAIL PROTECTED] Subject: Re: What events cause a media image to be taken? Michael, The Backup and Recovery Redbook says that, in addition to the times you mention, Media Images are taken when a queue becomes empty. The URL is: http://www.redbooks.ibm.com/redbooks/SG245222.html <http://www.redbooks.ibm.com/redbooks/SG245222.html> ...and you want to see Section 1.3 for the detail. This explains why the log file advances at times. You should be able to duplicate this on the laptop. We also had a problem with manually moving the log file if Permanent Dynamic queues were present. I believe this bug has been fixed in the latest CSD. Finally, we noticed a difference in the list of objects backed up depending on whether we did "\*" or "all". Not sure if this has been fixed. If you look at Brian Shelden's Log File Maintenance Support Pack (I forget the number), he uses "all" in his scripts for this reason. -- T.Rob -----Original Message----- From: Michael F Murphy/AZ/US/MQSolutions [mailto:[EMAIL PROTECTED]] Sent: Thursday, December 12, 2002 12:22 AM To: [EMAIL PROTECTED] Subject: What events cause a media image to be taken? I am getting deep into how a media image is taken, besides using rcdmqimg. The book simply states there are circumstances when a media image is taken, but it doesn't say what those are. I have been testing ways a media image is taken and when a checkpoint is taken. I am doing this on 5.3 on Windows but I think it should apply to unix and previous versions too. I have been doing a simple test. I put thousands of persistent messages on a queue and take them off at the same time. I can see the AMQ7467 and AMQ7468 written at regular intervals and the last log to restart the queue manager is moving forward fine. However, the log needed for media recovery never moves. I know you are supposed to take media images, but bear with me and I'll explain why I am doing this. In my tests, I take a media image with rcdmqimg -m QmgrName -t * * and then check the logs, no new messages. I continue puts and gets and eventually I see the log needed for media recovery moves forward the next time it is written. I assume these messages are written at each checkpoint. Now my issue.... I see queue managers where the media image is taken with these two commands: rcdmqimg -m QmgrName -t ctlg rcdmqimg -m QmgrName -t syncfile Since this is not all the objects, I would think the log needed for media recovery would never move, but in production it does. So I test this in the same way as my other tests. When I test it, the log needed for media recovery never moves. It only moves if I record all objects. The question is how is the log needed for media recovery moving forward on the production machines if record media image is not being done on all objects? Am I misunderstanding something? My understanding is media images are taken manually and during shutdown. The other interesting thing is that I see the log needed for media recovery move forward between the times that a media image is taken. This tells me some other process is taking media images, not just rcdmqimg. So what is taking the media images of all these objects? I need to know because from my observation, my client is taking media images on a schedule for nothing. In fact, it isn't actually doing anything. This does no harm on the high volume queue managers but the low volume ones get full filesystems because you can't purge any logs. This leads me to think that it has something to do with volume. I can't push the same volume through my little laptop that goes through the production servers so I can't duplicate this. Anyone? Anyone? Mike Murphy Sr. Middleware Consultant MQ Solutions, LLC http://www.mqsolutions.com Instructions for managing your mailing list subscription are provided in the Listserv General Users Guide available at http://www.lsoft.com Archive: http://vm.akh-wien.ac.at/MQSeries.archive Instructions for managing your mailing list subscription are provided in the Listserv General Users Guide available at http://www.lsoft.com Archive: http://vm.akh-wien.ac.at/MQSeries.archive
