I would mirror Jeff's responses here, especially #1. I am going to be heavily using the new scheduled discoveries, and would like the ability to omit.
I have to also ask about number 2. I haven't changed my network at all and have an SL500 NPM license... now when I run network discovery it discovers everything over again even though they are already in the database... then when I try to get past adding the topology information discovered to the database it kicks out an error saying my license does cover this many elements. This is really frustrating because it's double counting all my elements and won't let me get the topology data added to my database so "connect now" will work. I would have only expected the banner to list the real new nodes found... not all of the ones already in the database.
So basically it looks like this:
If you want Network Sonar to work you better buy a license for twice as many elements as you have or it WILL NOT WORK!
We have a known bug where the number of nodes reported as newly discovered in the banner is incorrect. The dev team is working to resolve this one...
Brandon has helped me to get though the running of network sonar without violating my licensed number of elements. After running network sonar during the import process I was able to select a few devices without all their interfaces and have them then checked against the database. At this point Sonar realized the node or nodes were already in my database and didn't re-add them. I split up my total devices already in the NPM database into a couple groups and then loaded them from the ran discovery profile one group at a time. This worked so if you too are running into a similar problem with a too many elements violating your element limit... take a subset of devices and only select some of the interfaces to test against the NPM database if they are already in NPM.
This didn't get my topology to show under the Network Topology resource but this could be due to the fact that my network requires cdp NOT be enabled. I did try running the following query as a work around:
INSERT INTO DiscoveryProfileOptions
SELECT Distinct(ProfileID), 15 AS OptionID FROM DiscoveryProfileOptions WHERE ProfileID NOT IN (
SELECT Distinct(ProfileID) FROM DiscoveryProfileOptions
WHERE 15 = OptionID)
I will experiment some more today and see if there's anything I've missed.
EDIT: Props to Brandon for being very active in tracking down issues that crop up with v10.
9) Ability to edit more than one discovery at a time. Initially I set all of my discoveries to run every hour. I think I've overloaded my server by doing this, so I'm backing off and going try every 6 hours. It would be nice to be able to select more than one discovery and change the run time on all of them.
One more thing, I think this has been mentioned, but I would like to be able to select the actual time that each discovery runs. There would be numerous reaons for this that I can think of.
I could be wrong but with #2 it looks like after Network Sonar runs it totals up the discovered elements and then adds this number to the total number of elements already in your database and then checks this total against your license. The problem is it does this before ever checking to see if they are in the database already ie. a duplicate. For me at least this immediately caused a license issue. As you can read in my above post I did find a way around this with Brandon's help by not selecting all the interface or device at the same time.
My .02 from jeff's list...
1. Absolutely. On the screen Device Types to Import, I'd like an ignore button on this screen. Agree on interface types.
2. I see Brandon says this is being worked on.
4. I have disappeared interfaces. This is new to 10RC. Never had these before.
5. I'd rather have the Discovery screen show the date and time of the next discovery. I don't run all scans normally.
6. I set my schedule for 168 hours, but I'd like them to start on different days or at specific times. We can't set that.
7. I ended up setting the timeout to 240 minutes as discovery was timing out on a /16 subnet before topology was discovered.
8. I set hops to 1 in hopes of better topology discovery. Not sure this helped.