Antw: Re: finding inplace alters
Hallo, in Version 7.31 the table syssltdatt is not present in sysmaster. Greetings Heinz >>> Fernando Nunes <domusonline@gmail.com> 22.12.2010 01:47 >>> On Tue, Dec 21, 2010 at 6:37 PM, Mark Scranton <mark@markscranton.com>wrote: > On Dec 20, 4:22 pm, PeterP <peterp...@gmail.com> wrote: > > On Dec 18, 10:52 pm, Art Kagel <art.ka...@gmail.com> wrote: > > > > > The IPA reversion problem is fixed as of 11.70. No longer a problem. > > > > > Art > > > > > Art S. Kagel > > > Advanced DataTools (www.advancedatatools.com) > > > IIUG Board of Directors (a...@iiug.org) > > > Blog:http://informix-myview.blogspot.com/ > > > > I thought it was fixed in 11.50 with IPA pages now being treated > > differently since 11.50 or was it just the rollback code for revision. > > I can't easily dd out the partition pages at the moment and guess / > > work out what's changed in the structures. > > 11.70 just tells you how many pages have IPA for performance I > > thought. Got any links ..? > > > > Still there are IPA bugs for things like compression etc. > > Generally I don't worry about them (unless you have almost 255) but > > everytime a table has a problem it's another thing to consider, > > especially with fragmented. > > > > You would think there would be a tbl_IPA_cleaner thread or at least > > some kind of IPA admin fix task. > > Maybe a environment variable like NOFUZZYCKPT (strangely similar to > > IPA) NOIPA for when you DDL small tables. > > Anyone else had sleepless nights over damn IPA's with their associated > > LTX's. > > Yeah, Onion was the tool I was thinking of. )) > > I've had a script forever that identifies IPA's, but am certain it > need tweaked to be complete - even after you resolve the IPAs, the > script will still identify them. I believe that Jonathon (Leffler) > actually modified the code to take care of this issue. I will post it > here in a few minutes. For the record, what you're looking for are > "slot 6 pages", or as Andreas said, "secondary partition pages." > Traditionally, each partition has a single partition page that has 5 > slots. With an IPA there is another page, the slot 6 page. My biggest > usage of the IPA script was in fact the reversion issue, followed by > any potential performance concerns. As Art mentioned, the reversion > issue was fixed in 11.7 which is a huge and important fix. For those > of you not on 11.7 (many I would think for now), word on the street > was always "you can't upgrade if there are IPAs." That was always a > myth - you can't revert (easily) if there are IPAs. I used to mention > this "all over the world", and for the longest time very few people > ever believed it. If they ever "hit it", then they believed it. ;) > > Thanks - > Mark Scranton > The Mark Scranton Group,LLC > _______________________________________________ > Informix-list mailing list > Informix-list@iiug.org > http://www.iiug.org/mailman/listinfo/informix-list > I believe it's possible to get the pending in place alters using SQL. But it needs syssltdat and I'm not sure when it was introduced. And yes, as Andreas wrote this requires some knowledge about the internal structures... but then again... You can't get more internal than Onion :) We can get the extending partition pages from sysmaster, and then get the slot 6 data from them and interpret it to get the number of pages in each version. If any returns != 0 we have a pending inplace alter. This should run fairly quickly but I'd need to check this... If anybody has a V10 or V9 or v7 at hand I'd like to know if syssltdatt is present in sysmaster -- Fernando Nunes Portugal http://informix-technology.blogspot.com My email works... but I don't check it frequently... -- WESTFLEISCH eG * Hauptsitz: Brockhoffstr. 11, 48143 Münster Amtsgericht Münster: Gen.-Reg. 307 Aufsichtsratsvorsitzender: Heinz Westkämper Vorstand: Dirk Niederstucke, Peter Piekenbrock, Josef Lehmenküh ler, Dr. Bernd Cordes, Dr. Helfried Giesen Hinweise: Es können nur Mails bis 30 MB empfangen werden. PowerPoint-Dateien müssen in eine ZIP-Datei gepackt werden. --------------------------------------------------