observations when testing V11.50.UC4

Hi experts, when creating a dbspace to hold the logical logs and adding some logical logs using onparams -a I see that a checkpoint is triggered after each piece added. Now some ghost in the machine records a checkpoint every nanosecond and therefore keeps shouting at me, that my logical log pieces are to short.... Fact is: I know what I am doing. Can I switch off the unwanted and unneccessary checkpoints after each and every onparams -a?Somehow I do have a deja vu of 'self healing & Co'. All I want is to add 1024 llog pieces then do a checkpoint. I hope this does not have to be a feature request! Thru versions 7.X and upto 10.00 this never was a problem. Another thingie is, that when having dropped all the logical log pieces in the dbspace I am moving them out from, a fake backup (ontape ... -F) does not do the job anymore. They still hanging around in pre-drop mode ('D'). I then did an ontape -s ... not using '-F' to the bitbucket to get rid of them. Now this is funny in the labs having 10MB of logical logs and not so funny with databases having 50 GB worth of logical logs and raising TX/sec rates. I mean I do not test this w/o reason. To move logical logs into a new dbspace and resizing the pieces at the same time is a very easy and common thing to do, isn't it? Is there a more modern style procedure to move out and resize the logical log, which I missed? If YES: Where is it in the docs? If NO: Why is there new annoying stuff happening in trivial areas? Sorry for the rant! dic_k -- Richard Kofler SOLID STATE EDV Dienstleistungen GmbH Vienna/Austria/Europe

I find those messages annoying, too. And I do not know how to turn them off. But just for grins, make each logical log bigger until the messages go away. It might be a size that you can live with.

LIGHT SCANS schrieb: > I find those messages annoying, too. And I do not know how to turn > them off. But just for grins, make each logical log bigger until the > messages go away. It might be a size that you can live with. Hi LSC, well, no! Just let me repeat some of what I posted: I need to change the size per logical log piece from 75000 2KBpages to 100000 2KBpages because peak TX rate went up by 15% during last quarter and log switch time went down from 5 mins to almost 4 mins. So I created a new chunk to hold the new & bigger logical log pieces. Much to my surprise I get a checkpoint after each onparams -a. This is like 8 chkpoints / second.I never noticed these checkpoints in IDS V10 or earlier. Some ghost in the machine seems to take notice of very frequent checkpointing and starts shouting at me telling me that the size of the llog pieces is too small. So of course for 1 test I tried using 500000 2KBpages (~ 1GB) for the logical log pieces -> no change in behaviour... Therefore I wrote, that I actually want to be able to switch of the pointless checkpointing and do excatly 1 checkpoint manually when all the llog pieces are there. I did not test to switch off one new member of the 'Selfhealing & Co family', the new AUTO_CKPTS which defaults to 1 (=ON). But this is one of the next things on my testing list. In my opinion on a huge & busy system it is a very good idea to switch off all the AUTO_* envvars - which all default to 1 (=ON) - until they are fully understood & tested! dic_k -- Richard Kofler SOLID STATE EDV Dienstleistungen GmbH Vienna/Austria/Europe