Re: General IDS 11 question

----- Original Message ----- From: "John Carlson" <jwcarlson1@yahoo.com.invalid> Newsgroups: comp.databases.informix To: <informix-list@iiug.org> Sent: Wednesday, July 25, 2007 10:13 PM Subject: General IDS 11 question > Of those of you who are using IDS 11 in a production environment . . . . > > How is IDS 11 working for your application? > Any issues / gotchas / glaring bugs that you've had to deal with? > Good performance? > > Just curious . . . thanks in advance. > > JWC > _______________________________________________ > Informix-list mailing list > Informix-list@iiug.org > http://www.iiug.org/mailman/listinfo/informix-list Just recently moved our server performance statistics collection database off RHEL3/10.00.UC5 to RHEL4/11.10.UC1 in the last few days. As expected the installation/migration went smooth and of course the applications that worked with 10.0 work with 11.1 with no modifications. I feel kind of dirty setting LRU_MIN/LRU_MAX so high (using the defaults of 50/60) but I want to take the new fangled non blocking checkpoints and autonomic features for a spin with all AUTO_* paramters set and RTO_SERVER_RESTART set to 180 (probably a bit high). It is really really really weird not seeing a checkpoint fire every 5 minutes and only when it is needed to maintain the recovery time objective. Still waiting for my first system generated checkpoint. With my large number of buffers, high lru min/max and a whopping 288 insert/update/deletes per hour I still don't have a LRU write. My LRU queues are currently 20% dirty. It might take a while to reach 50% dirty. The sql history feature is very cool, this will come in very handy when we migrate the real applications to 11.1, especially since we are a smaller shop and there is a good amount of ad hoc dbaccessing going on. DIRECT_IO to cooked chunks is on and working just fine. I got my first performance advisory in the online.log today. I was doing a larger than normal delete of some data and got a warning about logical log size being too small because it filled in less than 30 seconds. 12:15:55 Performance Advisory: Logical log file size might be too small fora checkpoint to complete. 12:15:55 Results: The size of individual logical log files is too smallfor the current workload, resulting in each log file filling very quickly. If log files fill in less than 30 seconds, the checkpoint might remain blocked because the last log file fills during the time needed to perform the checkpoint. 12:15:55 Action: Increase the size of the individual logical log files so that it takes at least 30 seconds to fill each one. Look at the online log to determine how quickly the log files are filling, and then increase the size of the files proportionately. Performance seems faster. This engine stores statistics supplied by the other instances in our network and spits them back to a php application that displays this information graphically. The graphs seem to take less time to render running on older hardware and cooked chunks to an internal disk. Of course this could very well be a result of getting this data into a dedicated server where it does not have to fight other applications for resources. No bugs or problems to report so far. 100% uptime. Improvements to onstat thread output are very cool. It is nice to get more information about my threads instead of the old sleeping forever song and dance. New information for the cleaner threads shows the number of wakeups and idle time for each thread. Neat. No HDR, ER, RSS or SDS for this guy yet, so I can't comment on those features. I rebuilt some indexes and am happy to report that statistics and distributions were automatically collected during the index build. Very nice. DD_HASHSIZE, DD_HASHMAX, DS_HASHSIZE, DS_POOLSIZE and BTSCANNER are still not included in onconfig.std. I am not sure why. That is about all I can think of right now, it has only been 2 days. Andrew

> I feel kind of dirty setting LRU_MIN/LRU_MAX so high... > Andrew- Hide quoted text - > > - Show quoted text - Oh, Art must be spinning in his grave right about now... :) -Nate

On Jul 27, 1:20 am, nate...@gmail.com wrote: > > I feel kind of dirty setting LRU_MIN/LRU_MAX so high... > > Andrew- Hide quoted text - > > > - Show quoted text - > > Oh, Art must be spinning in his grave right about now... :) Didn't even know I was dead yet! Naw, I can live with that as long as check points no long block I'm happy. Art S. Kagel

On Jul 27, 4:31 am, "Art S. Kagel" <art.ka...@gmail.com> wrote: > On Jul 27, 1:20 am, nate...@gmail.com wrote: > > > > I feel kind of dirty setting LRU_MIN/LRU_MAX so high... > > > Andrew- Hide quoted text - > > > > - Show quoted text - > > > Oh, Art must be spinning in his grave right about now... :) > > Didn't even know I was dead yet! > > Naw, I can live with that as long as check points no long block I'm > happy. > > Art S. Kagel Heh, just my poor attempt at humor. :)