GaryP

Hi, just to add a little more detail to this: We're seeing the AUTHFAILs in clumps of 4-6 messages, triggered approximately 60s apart, and always between midnight and roughly 1:40am Very curious....

Hi all, I just posted a rather complex question regarding alerting based on Trap Detail over here: Hopefully the functionality to search Trap Detail will make it into 10.0 I'd also really like to see the ability to create and assign custom scales to graphs, this is really high on my list of priorities as I monitor a lot of…

Yes, we see exactly the same on our system. '[Unknown]' are devices that are only ICMP monitored and hence have no SNMP data whereas 'Unknown' are SNMP monitored but do not have a corresponding entry in the sysobjectid database.

Cheers Brandon, can you be more specific as to what diagnostics you require?

Cheers Byron, looking forward to this getting fixed...

Cheers Chris, I realise the difference between the monitor and the log forwarder and it was our server guys who specifically requested the exclude functionality. I'll ask them if they can write up a more specific case for me, hopefully it may make a little more sense then (to you and me both!).

Yeah, it's also worth noting that I can't differentiate between messages coming from different nodes with the current functionality, either. If I set up a threshold for PORT_SECURITY-2-PSECURE_VIOLATION, for example, that triggers after 10 messages it can't differentiate between 8 messages from one host and 2 from another.

Hi Miron, I think it basically needs to go through a full poll before the manage/unmanage icons are properly reflected after each change. I managed to get the interfaces to be polled properly again following advice from here:

Cheers Mark, that only really changes the colour scheme of the layout elements of the pages though, rather than the data visualisations themselves. I'd really like to be able to change the colour schemes used by the graphs and the reports.

Thanks Matthew, I've posted the details over there but almost all of the devices are already in the powernet.mib and aren't coming up as 'Unknown' in Orion NPM. The problem is that they're reporting the same information for Vendor and Machine Type so Orion NPM can't tell what hardware they are.

Grr...just been and tested this and you're mostly correct. I'm primarily a Mac user (with Safari 4.0.4 under OS X 10.6.2) and although my system knows it's in the UK and is set to use UK date, time and number format, Orion's Web console doesn't reflect this. Same goes if I'm using Chrome or Firefox on the same machine.…

Gamatronic GMaCi Power+ 1.3.6.1.4.1.6050.5 It's a UPS monitoring device.

Hi, I've got another thread open ( ) where I was recommended to post in this thread. I don't actually think this is going to help as these devices aren't being displayed as 'Unknown' machine type or vendor but, instead, are misreporting their machine type as 'American Power Conversion Corp.' and some of the NetBotz are…

Thanks Denny, I've done that now and it's being worked on. I'll post any resolution here for the benefit of the community.

Cheers bshopp, I'll have to modify my bulk deletion scripts to take this into account as I fear I may now have rather a lot of historical data left cluttering up the database.

I have a similar network architecture but approached the problem in a different way. I chose to monitor all switch ports on all devices (just to have traffic stats), however I marked all the ports on the user switches as 'unpluggable'. It is fairly simple with a little SQL to mark all the interfaces of a given node…

All of these interaces have an index which can be confusing because, from my experience, each time a new module inserted in say a Cisco router... the interfaces are renumbered ie. their index might be different now. Try this magic sauce out on your routers , I believe it'll sort you out in this respect (I know this works…

could you run a discovery covering the IP range of the equipment involved and during the import phase select the relevant device types and interface types that you are interested in. This would add all the interfaces into the database. You could then set all those which you do not want stats for to the "Unmonitored" state…

Cheers Michal. Seems I was using an old version of Firefox to do my testing with.

Cheers Marie, that'd be super useful!

This is very interesting, we've also been seeing SNMP-3-AUTHFAIL syslogs sporadically appearing from some of our monitored devices with the offending address being it's poller.

Thanks for this, it will be very useful in killing off our Nagios install I can't seem to get the report to sort by LastTimeUp, though. I'm gonna have a hack about and see if I can sort it.

You can't usually go wrong with O'Reilly: oreilly.com/.../9780596000202 Or, if you're in the UK, I can highly recommend Kenson's Solarwinds training courses.

There was a bit of a discussion going on over in the SCP sub-forum about certified members being given very limited license copies of software. This is exactly the sort of situation they would be useful in.

Do you have your web console's IIS site modified to only accept SSL connections? Ours is and Configuration Wizard always fails at the same point for us, too.

Hi there, this isn't a solution to your exact query but it works for us: We've been monitoring link utilisation between our core routers on campus for a few years using an open source app called Network Weathermap (http://www.network-weathermap.com/). It's a visualisation tool that takes existing data (from sources such as…

Hi cereal83, I had exactly the same problem scanning our network. We've got 1000+ Cisco devices (spread across two class Bs and a couple of class Cs) and a full network scan was only returning 45 devices. Solarwinds support had me jumping through hoops for weeks trying all sorts of different things (manually specifying…

The actual content of the trap itself never uses hostnames, only IP addresses. In addition to sending out the body text of the trap in your email alert you could try including ${HOSTNAME} or ${CAPTION} to get a more meaningful identification of your devices. Orion NPM will interpret those variables from it's own database…

Hi, just wanted to add a 'me too' to this one. We've disabled passphrase logins on our servers for security reasons and only use keypairs. It would be great to see this functionality in APM.