I've seen it in unicenter, I don't remember CA's vernacular for it though. Ultimately you set your default thresholds, then after some period of time (like a week or something), you manually triage any alerts that the nms generated pertaining to the device.
You then decide manually which ones are real and which ones you would consider false alarms. This helps to alleviate false-positives after go-live date which often cause unjustified panic amongst people unable to triage on their own - read: customers. After go-live, the NMS will offer to "learn about" any alerts that it generates. An example given earlier in this thread about dropped packet rates being proportional to total throughput is a good example - you could potentially either ignore, or modify thresholds during a timeframe. So if the NMS generates an alert for discards, you would enter a menu on that alert where it would give you a set of options. Options including, permanently ignore this alert or setup a maintenance window for this alert where during this timeframe, exact duplicates of this alert are ignored but others come through as normal. The benefits of this increase dramatically of course as you noted, when coupled with the scheduled maintenance window (either one-time or rolling). How to solve unplanned reboots ? That's where the manual triage process would come in. I think that is a matter of sysadmin policy, where you would never allow your NMS to be taught to ignore such things that occur outside of maintenance windows. Best Regards, Jeremy Eder, RHCE, VCP -----Original Message----- From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf Of Randy Millsop Sent: Monday, December 01, 2008 12:44 PM To: InterMapper Discussion Subject: Re: [IM-Talk] Feature for XMAS Yes, the 'scheduled maintenance' ack feature would be greatly appreciated. We have both the need for one-time schedules and recurring schedules. The recurring schedules vary from daily, every other day, weekly/monthly. Jeremy, the concept of the learning threshold is very intriguing. How have you seen this being used, and by which software? I haven't seen this in the real world, so I can't be sure it'd solve some of our issues or not. For instance, the varying reboot time frames of different types of devices. We certainly wouldn't want to miss unplanned reboots, but used in conjunction with the scheduled maintenance windows this could be useful. Thanks, Randy Millsop Network Administrator San Joaquin Delta College ----- Original Message ----- From: "Andrey Gordon" <[EMAIL PROTECTED]> To: "InterMapper Discussion" <[email protected]> Sent: Monday, December 1, 2008 9:28:22 AM GMT -08:00 US/Canada Pacific Subject: Re: [IM-Talk] Feature for XMAS I will second the scheduled maintenance window feature request. It's something that I've been asking for for over a year now. I see it (as it is implemented in just about any other NMS) just as an ACK from TIME and DATE to TIME and DATE. __________________________________________________________ Andrey Gordon | Integrity Interactive | Network Engineer | +1.781.398.3518 > From: Jeremy Eder <[EMAIL PROTECTED]> > Reply-To: InterMapper Discussion <[email protected]> > Date: Mon, 1 Dec 2008 12:21:01 -0500 > To: InterMapper Discussion <[email protected]> > Subject: RE: [IM-Talk] Feature for XMAS > > The concepts of: > > - scheduled maintenance windows (recurring or one-time) per-device and/or per > map (not the same as timed-ack's) > - a concept I call learning-thresholds (auto-adjusting during a learning > timeframe, and thereafter become hard-set..other nms have this great feature) > > are 2 that we would like (in priority order). > > > > > > Best Regards, > > Jeremy Eder, RHCE, VCP > > > > -----Original Message----- > From: [email protected] > [mailto:[EMAIL PROTECTED] On Behalf Of jeff donovan > Sent: Wednesday, November 26, 2008 10:39 AM > To: InterMapper Discussion > Subject: Re: [IM-Talk] Feature for XMAS > > > On Nov 26, 2008, at 10:26 AM, Bernd Janke-Wohltmann wrote: > >> Hello together, >> >> >> >> here is my reseller X-MAS list of wisches: >> >> >> >> - Expand the Interface Windows of switches/Routers for per >> Port: >> >> o Ignore Outage >> >> o Ignore Discards >> >> o Ignore Errors >> >> - Active hours for Devices >> >> o When a device fails during Inactive hours this is like ignore >> outage >> during this time >> >> - ACK maps status >> >> o Compute a HASH about MAP status >> >> o If the hash changes - then there is a new status in the submap >> >> >> >> The goal is to keep our cascade maps green for normal operation. >> >> >> >> Greetings from the cold and gray north of Germany >> >> Ciao, > > Greetings Ciao, > I love your ideas. i have been running 4.2 for years and have switched > over to 5.0.5 and have noticed some of the same needs that you have. > > Maps that once would show alarm, ( orange ) are now ( Red ). > > I would to make that color change / Acknowledge that Map status or > device. or set the level somehow. set alarm threshold,.. possible ? > -:) > > -j____________________________________________________________________ > List archives: > http://www.mail-archive.com/intermapper-talk%40list.dartware.com/ > To unsubscribe: send email to: [EMAIL PROTECTED] > > ____________________________________________________________________ > List archives: > http://www.mail-archive.com/intermapper-talk%40list.dartware.com/ > To unsubscribe: send email to: [EMAIL PROTECTED] > ____________________________________________________________________ List archives: http://www.mail-archive.com/intermapper-talk%40list.dartware.com/ To unsubscribe: send email to: [EMAIL PROTECTED] ____________________________________________________________________ List archives: http://www.mail-archive.com/intermapper-talk%40list.dartware.com/ To unsubscribe: send email to: [EMAIL PROTECTED]
