sedmo

It looks to me like you could do this by selecting Interface for the Type of Property to Monitor. This appears to give access to both Node and Interface fields.

There are two name fields in the solarwinds database, the NodeName and the SysName. The SysName is acquired through snmp from the device being monitored. The NodeName is the name you configure within System Manager. It sounds like your alert is using the SysName field? If so, you need to change your alert to use the…

I think a filter like this should get you what you are looking for.

The answer to each of your questions about why the software does not work the way you would prefer it to is because it's not currently designed and built that way. If a product does not meet the criteria outlined in your evaluation process then you should probably select a different product.

I believe you need to copy the orionreportwriter.schema file to your separate web server

Try this. Modify this line in your query: eventtime between dateadd(week, -1, getdate()) and getdate() Change it to look like this: eventtime between dateadd(month, -6, getdate()) and getdate()

Use Nodes.StatusDescription in your SQL instead of Nodes.Status. Status contains a numeric representation of the node status and StatusDescription is a text description of the node status. As for your second question about your alert not triggering correctly, it would be easier to troubleshoot if you could post the details…

You may also want to look into TACACS accounting if you are using TACACS for centralized authentication.

tsperbeck, You can't adjust the interval between the fast polls but you can adjust how long NPM performs fast polls before marking the node as down. In version 9.1 this is configured in System Manager through File ==> Advanced Settings ==> Node Warning Interval tab. I'm not sure if this path is the same in 9.5.

The numbers represent the number of days since 1/1/1900. You can experiment with this by executing the SQL listed below.DECLARE @StartDate DateTime DECLARE @EndDate DateTimeSET @StartDate = 38655 SET @EndDate = 38664 select @StartDate as Start, @EndDate as Finish This should return: Start Finish…

You need to stop the solarwinds services before editing the database directly in order to be sure that your changes will be permanent. Sometimes information from the database tables is held in memory and can be written back to the database.

You should be able to do this without using a regular expression. On the Message tab of the syslog rule enter the ip address you want to trigger on in the Syslog Message Pattern box. This will look for the ip address in the message part of the syslog message. The configuration would look like this: Then on the Trigger…

I think what you are looking for is a WHERE clause. SELECT Caption,VendorIcon,Ip_Address, DateDiff(hour,Nodes.LastBoot,GetDate()) AS HoursUp FROM Nodes WHERE Location = 'usa' ORDER BY 1 ASC

See if the filter criteria in the screenshot below does what you are looking for. I'm interpreting your criteria as follows: Include all interfaces for node RDE-3640#1 Include only interfaces Fa2/0 and Fa0/1 for node RDE-3640#2 Include all interfaces for node RDE-C6513-1 Include only the interface that has interface index…

Have you tried modifying the data portion of the icmp packets used by the Solarwinds tools? See this thread for more info:

It doesn't look like the SNMPVersion field is available through report writer but you could create an Advanced SQL report to do this. A query like the one below would give you a very basic report select caption, snmpversion from nodes

I believe that means that something is trying to query this server with SNMP but is using an incorrect community string.

Is your NodeGroup field a custom property? 

I saw similar behavior with the 7600 series devices. What I found was the CPU utilization OID returned multiple values (a table) on these devices and NPM appeared to be averaging all of the values together. The first entry in the table contained the total cpu utilization number that I was looking for so I solved the…

Is the trigger condition set to ANY instead of ALL?

which field do you want to update?

It uses the InOctets and OutOctets counters.

I believe what you are looking for is in the CustomPollerStatus table.

I think your filter needs to look like this. Owner like 'servers*' or Owner = 'routers'

On the Filter Results tab of the report you should be able to enter criteria that looks like this: Records where NodeName starts with 1 -

I believe the problem is caused by using "equal" in your trigger condition for the node name. I think you need to change this to: Node Name contains Customer_

You can do this by setting up a rule with the Syslog Viewer app. Select Add New Rule, then on the message tab enter *printer* in the Syslog Message Pattern box. Then go to the Alert Actions tab and select an action to take for messages that match the rule.

This could be caused by Solarwinds not using 64 bit counters when polling the device.

It's very possible that the "is not empty" criteria is causing you problems. There are two different ways that database fields can be empty, they can be blank or they can be Null. when creating an SQL query to find blanks in a field you would use syntax like this (note that the quotes are two single quotes): Field1 = ''…

This may be happening because the data types of the two fields are not the same. The CPU Load field is an integer so your custom property field (CPULoadMaxPercent) must be integer also.