lhorstma

SELECT Distinct AlertStatus.TriggerTimeStamp, AlertStatus.ObjectName, AlertDefinitions.AlertName, AppNode.Caption AS ApplicationNode FROM AlertStatus WITH (NOLOCK) INNER JOIN ActionDefinitions ON AlertStatus.AlertDefID = ActionDefinitions.AlertDefID INNER JOIN AlertDefinitions ON AlertStatus.AlertDefID =…

I could, and that's basically what I'm doing, just manually in SQL. It would be nice however, to have the icon represent the state of the element just like with the interface icon. It would make it much easier on me to do my maintenance if I can visually see what's a problem right there on the page. I have thousands of…

I was just coming here to suggest the same thing. Also, the "Group Members" should have a "Group Members - Tree" alternative, like the "All Nodes - Tree", where you can group things together by up to three levels.

FYI, the 10.1.1 update did NOT fix any of my problems. Tickets 203615 and 204406. Last night the collector data processor service failed on 3 polling servers and volume statistics stopped making it to the database on 2 polling servers (with no other indication of issue).

huh? yokoheight? It's spam, I've reported it. As for the parent issue, what version of APM are you running? I'm on 3.5 and am going to check to see if I get the same thing as it looks like it's not reporting group status for applications for you. My understanding is APM 4 has better integration with the groups and may…

Mike, yes, it was using the "Find Processes, Services, and Performance Counters". The issue had spread to a whole host of other monitors over night. Over half of my performance counters had stopped working with the same "Bad input parameter" error, so I figured it had to be something with the Orion server itself. I found a…

I don't use basic alerts either. They don't span pollers so if you load balance devices from one poller to another you need to redo your alerts. The only instance I could see basic alerts being useful is on very small implementations, like if you were monitoring less than a hundred devices on just one poller.

I second this, I would find it very useful.

I guess i would like to be able to look back during the event to see what part of this logic didn't work. Then I started thinking, can I reference Node custom properties even though I have the "Type of Property to Monitor" set to Interface? It should work as long as you don't change the "Type of Property to Monitor".…

While I agree that it would be nice to have this, and especially for anyone newer to the advanced alerts system, I wanted to mention that you can already do this with just one extra step. Add a new condition group and mark it as "Trigger Alert when NONE of the following apply", then create a new simple condition under that…

I created a template using the TCP port monitor to check port 135, and then created alerts on that to send an alert if the node is up but RPC is down. So if the server responds to pings but not to RPC I get an alert. We mainly get alerts for servers when they hang on reboot, or general RPC failures, but do get a few false…

To add to these suggestions I would like to see some additional filtering tools added to the discoveries. One would be adding interface types to the ignore list, so instead of adding each individual VOIP interface to the ignore list for every router I could add a single entry to automatically ignore all interfaces named…

You just want a "does not contain" right? That's an easy one, just add a none condition group: Trigger Alert when all of the following apply ..custom-property1 is equal to toilet ..interface status is equal to down ..Trigger Alert when none of the following apply ....interface fullname contains bri Or if you want to…

I created different alerts for each alert level rather than trying to create one alert for multiple levels. We also separated them by the size of the disk, so a drive that's 1 TB would alert at 50GB, 25GB, then 3GB left, but a 500GB drive would alert at 20GB, 10GB, and 3GB. All in all we have a ton of alerts set up and…

I don't have one already written up for that purpose, but it shouldn't be too tough to write a vbscript to use psexec to run nmap against the target and parse the result. If I get some free time I'll see what I can come up with.

1) How many alerts of ActionType Email fired in the previous 24 hours, 48, 72, day/week/month/year/range? Of those, how many triggered emails? What addresses were they sent to? I was playing with something similar earlier today. I wrote a report in Report Writer to give me alert log info. Basically Management asked why an…

Well I wrote my own and got it working. Pasted here for anyone else to use or modify. Just change the tnsname to what you want to ping, the temporary output file to something unique, and check the path to your tnsping. Returns status of up if it get's an OK result, and down if it gets anything else, plus the time in msecs…

The alerts create the events as a trigger action (see below). You can alter or disable the associated alerts. Instead of disabling the alert I would probably set under trigger condition the "Do not trigger this action until condition exists for more than" and set it for 6 minutes or so so that it must be active for at…

Here's an idea which you might check that I've had an issue with before; Under Trigger condition, make sure "Type of Property to Monitor" was the same for every condition listed. My previous issue appeared when selecting a "Node Name" condition using the Property to monitor: Node, and then switching to Property to monitor:…

I haven't tested this but can you do the "Define SQL Variable" part of select variable and put the math directly in there? Would come out to be ${SQL:Select 100 - ${VolumePercentUsed}}. No time to test right now though.

Wow. I'm on 2k3 and sql 2005. Orion NPM 10.1.1 sp1, APM 3.5, and just added NTA 3.7. My retention time is close to default. NPM polling is 2min/10min. Also have 5 min polling for APM. Over 4 polling servers I have about 20,000 NPM elements, and though I don't know where I can see the exact count, about 3000-4000 APM…

You might want to check the net-snmp version and update it. I had a whole bunch of Redhat servers which had SNMP configured properly and would pass the SNMP test but would show you absolutely nothing when you moved to listing the resources. Updated the net-snmp package and they started working fine. I don't remember what…

I have a similar problem that I posted here about. The alert emails I'm getting are being sent in HTML instead of plain text now, which my ticketing system also now truncates (haven't counted the characters it cuts it off at yet though). Nobody has been able to give me any ideas on workarounds yet.

it stripped out my pic, lets try that again.

Seconded! That's a good one.

My 3 pollers and my database are all Xen guests and are running great (about 15,000 elements right now). I was worried when we first switched over from physical boxes but it's been very smooth.

I just finished my upgrade install, was told this would update would address my issue with the collector data processor service failing and also with losing Volume polling data. Upgrade went well, no errors given, everything appears up and running so far. I'll post any issues that appear/are solved that I come across with…

I second this. I need to be able to set this value to an alias instead of the primary poller.

I have a similar issue, the response I got from support is that I have a bad alert set up. They said: Open the \Program Files\Solarwinds\Orion\AlertingEngine.exe.config file.Edit the below section: < logger name="UpdateResetRows" > < additivity value="true" / > < level value="NONE" / > < / logger > Set the level value…

Shouldn't it be ${Nodes.City} with an "s"? Not positive it would work anyways but it seems that any time you need to display variables outside the alert context you just need to preface it with the table name, which is Nodes.