On Fri, 2010-03-12 at 13:28 -0500, mcclainje wrote: > Anyone using this? Any helpful advice? We are seeing quite a bit of > variation in the numbers and are not sure if we should be concerned.
We have created that probe primarily as a convenient means of tracking InterMapper performance on customer machines in the event of a performance problem; you are, of course, welcome to use it for your own purposes, but you should be aware that it remains in the Experimental section for a reason (we only promise that it will change, and in ways that probably aren't backwards compatible). The numbers that are in there at present are mostly there because we're working on things related to them. The engine frequency is the canary in the coal mine; if it remains near 66 Hz, everything else is probably fine with the server. If that number drops substantially (below 30Hz), you may see evidence of impaired performance; the other numbers might help us to pinpoint the cause. There are two pieces of advice that might be helpful when using it; one piece of advice is not to put a notifier on it; in our testing, most operating systems have some kind of maintenance routines that they go through every night which adversely affects InterMapper's performance even under otherwise optimal conditions (if you're using the database, the nightly database maintenance is one such task). As a result, you'll almost certainly get a notification in the wee hours of the morning every day. The other piece of advice is that you should only have a single instance of a device using that probe against a particular server. The numbers reported are all "since the last poll" numbers, and if you have two devices polling the same InterMapper server, they will share data and the numbers that you see will be difficult to interpret properly. The numbers will also change in meaning if you change the poll interval because of this. As an example, one way you might see quite a bit of variation is if you have enabled the database server, which pulls data from InterMapper once a minute, but the engine status probe is set to poll once every 30 seconds; during the 30 seconds around the top of a minute, you would see a lot more activity while InterMapper responds to that request for data than you will during the 30 seconds around the bottom of a minute. -- Stephen Ryan Dartware, LLC ____________________________________________________________________ List archives: http://www.mail-archive.com/intermapper-talk%40list.dartware.com/ To unsubscribe: send email to: [email protected]
