sedmo

The additional web installation will not interfere with the web site on the polling server.

It looks like there is a report named "Devices that do not respond to SNMP" under the Polling Status category.

I assumed that those interfaces where on the RI-NCC-2811 device. If that is not the case then the grouping will need to be modified. The way it is currently written, the interface has to match one of the five statements below. caption like RI-NCC-2811 and interfacename = Serial0/0/1 caption like RI-NCC-2811 and…

How about something like this? SELECT Nodes.Caption, Nodes.UnmanageFrom, Nodes.UnmanageUntil, Nodes.StatusDescription FROM Nodes where Unmanaged = 1 or unmanagefrom >= getdate() ORDER BY 1 ASC, 2 DESC

Great! Glad you got it working.

I believe the procedure updates a file called orionreportwriter.schema. Although the file name seems to indicate that it pertains to reports, it obviously is used by other parts of solarwinds.

Alfred, I think you will need to go with the alert that uses a suppression. There is an earlier post that has a screenshot example. 

I see the same. "none" translates to the usage of OR and "not all" translates to the usage of AND in the SQL syntax. I think however that the way the syntax gets configured is incorrect when there are multiple conditions under the "none" or "not all" conditions. For instance if I setup an alert that uses a "none" condition…

You'll find that dealing with dates in SQL can be very involved and there are many different ways of specifying the same thing. Here is what this particular query is doing. This line of the query finds the first day of the previous month. dateadd(mm,-1,dateadd(mm,datediff(mm,0,getdate()),0)) If you run just this line as…

Are you running NPM and your web server on the same box or does your web site run on a seperate server?

I think you have some problems with the groupings in your where statement. Try this. SELECT TOP 10000 Nodes.Caption AS NodeName, Interfaces.Caption AS Interface_Caption, AVG(Case InBandwidth When 0 Then 0 Else (In_Averagebps/InBandwidth) * 100 End) AS AVERAGE_of_Recv_Percent_Utilization, MAX(Case InBandwidth When 0 Then 0…

The polling completion stat shows up on the Polling Engine Status resource. This resource is listed under the Miscellaneous Resources.

It might be worthwhile to try updating the custom property schema. To do this: open the custom property editor right click on an empty part of the toolbar and select customize. select the commands tab. scroll down through the list of commands until you see update report schemas. drag the updte report schemas item up to an…

Looks like you didn't get the complete WHERE clause entered. Here is what the complete SQL should look like. SELECT TOP 10000 CONVERT(DateTime, LTRIM(MONTH(DateTime)) + '/01/' + LTRIM(YEAR(DateTime)), 101) AS SummaryMonth, Nodes.NodeID AS NodeID, Nodes.VendorIcon AS Vendor_Icon, Nodes.Caption AS NodeName, Nodes.IP_Address…

You may want to consider configuring a Loopback interface as Don suggested earlier in this thread and monitor the loopback ip address. Otherwise, your NPM is going to show the router as down if the ethernet interface goes down.

I was able to get this to work using the Total Bytes Transferred by Node - Last 7 Days report provided by Solarwinds and the filter Nodes.NodeID = ${NodeID} One possibility for your report showing blank would be if the node you are viewing is not included in the report. Is this possible?

I've seen times where specific alerts seem to get corrupted and simply recreating the alert from scratch causes it to work properly. If you haven't already tried this, I would recommend creating a new alert with the configuration above and see what happens. If that doesn't work then you might try configuring the alert…

LOL... That bug has bitten me a few times also ;)

The SQL syntax you used to test directly would not be the same syntax used in the alert email, so it seems possible that your issue could still be with the results being returned by the alert engine query. I would suggest testing the alert email with just the SQL to see if it returns the value you expect. This would…

Try this. SELECT Nodes.Caption AS NodeName, Events.EventTime AS Event_Time FROM nodes JOIN events on nodes.nodeid = events.networknode WHERE (EventTime BETWEEN dateadd(mm,-1,dateadd(mm,datediff(mm,0,getdate()),0)) AND dateadd(ms,-3,dateadd(mm,0,dateadd(mm,datediff(mm,0,getdate()),0))) ) AND ( (CONVERT(Char, EventTime,…

The way to make sure you are only using fields from the Events table is to only select fields from the Events category on the Select Fields tab of the report.

108 is a type code that determines the format of the time. In this case it returns HH:MM:SS. Would it be possible for you to post your complete query?

No problem. Always best to be cautious when making bulk changes. Also, if you run the first query posted above you will get a list of the nodes that will be changed by the update query.

One thing that could cause this problem is if you are selecting fields for the report from tables other than the events table. Another possibility could be something configured on the Time Frame tab that is causing records to be excluded.

The importance of using loopback IP's for monitoring may differ depending upon the size and topology of the network(s) you are monitoring. One problem with using the LAN interface IP for monitoring the device is that it can give you incorrect information about the availability of the device. If a router is up but the LAN…

Do you have any reports that use custom properties? If so, do they run ok?

One thing to note about the query. With the way the DateTime is currently coded, the report will always show availability data for February 2010. If you want the report to always show availability for "last month" based upon the current date then you will need to modify that part of your query. Here is one way you could…

You might take a look at the existing Availability reports. Perhaps you could modify one of them to get what you are looking for.

Sorry my previous post wasn't clearer. I was referring to the way the logic was being interpreted in the alert criteria shown in the screen shot posted above. Keep in mind that what you are doing with the GUI is creating an SQL query. For this alert the goal is to create a query that contains caption NOT like ABC in the…

Does anyone know exactly what the polling completion statistic represents or how it is calculated? If it is less than 100% does that mean that something is not getting polled? Do nodes that are down affect this statistic?