Version 1

    Hi All,

     

    As we all know that the in-build setup in Solarwinds for reboot works perfectly only for Windows systems, whereas when it comes to Linux systems the tool fails to provide us the right set of information and generally trigger false alarms.

    The reason behind it is that, the OID [.1.3.6.1.2.1.1.3.0 (Sysuptime)] which Solarwinds refers for fetching the last reboot information in Linux systems changes its value as n when there is a restart of SNMP service in them. As n when the SNMP service fails and restarts, the value provided by the Linux system on the OID to Solarwinds says that the system has rebooted, and hence Solarwinds trigger the alarm for the same, which is a false positive alarm.

    Since we cannot change the OID used by Solarwinds for fetching the right uptime, there is a work-around for the same using the SAM Application.

     

    By using the SAM Application, we can correctly configure the reboot alarms in Solarwinds.

     

    Pre-requisites:

    1. Server needs to be configured with SNMP (This is needed because without SNMP we won't be able to fetch any OID information from the server).

    2. You should be having SAM Module installed in your system.

     

    Process For SAM Configuration:

    1. Go to Application Manage Template > Create New Template > Add Component Monitors > Manually Add Component Monitors

    2. Select SNMP Monitor

    3. Fill the data in the below format:

    3.1 In this, the OID that we will provide is "1.3.6.1.2.1.25.1.1" which provides the information of HOST-RESOURCES-MIB::hrSystemUptime.

    3.2 The Custom SNMP port needs to be 0 in case we need to use the default port ( In case of any other port, kindly mention the same)

    3.3 The Statistic Threshold needs to be set at Critical with value Less Than 60000 (this is because, 60000 means 10 min, i.e as and when the server comes up and SAM polls the data, if the system uptime is less than 10 min, the status of application with change to critical) (This configuration is good for me because I am polling SAM application data every 5 min, it may vary with different environment).

    4. Create the template and start assigning the nodes to the same. (We can create a dynamic group and assign that group to this template so that manually not every linux server needs to be added in the template).

     

    Process For Alert Configuration:

    1. Go to manage alerts and create new alert.

    2. In trigger condition, use the below configuration:

    3. In alert action, you can perform any sort actions as per your requirement.

     

     

    Pros:

    1. Reboot alert will be exact and there will be no false positive alerts.

     

    Cons:

    1. In case of node down for Linux nodes, as n when the server comes up, in that case as well you will receive a reboot alert, because the uptime of the server will start from 0 itself, which will result in activating SAM Application providing that information as system reboot.

     

    I am still trying to get a way to remediate the same. As n when the solution is ready, the same will be updated.

     

    Regards,

    Ujjwal